Abstract
The high biological properties and low energy consumption of Spiking Neural Networks (SNNs) have brought much attention in recent years. However, the converted SNNs generally need large time steps to achieve satisfactory performance, which will result in high inference latency and computational resources increase. In this work, we propose a highly efficient and fast SNN for object detection. First, we build an initial compact ANN by using quantization training method of convolution layer fold batch normalization layer and neural network modification. Second, we theoretically analyze how to obtain the low complexity SNN correctly. Then, we propose a scale-aware pseudoquantization scheme to guarantee the correctness of the compact ANN to SNN. Third, we propose a continuous inference scheme by using a Feed-Forward Integrate-and-Fire (FewdIF) neuron to realize high-speed object detection. Experimental results show that our efficient SNN can achieve 118X speedup on GPU with only 1.5MB parameters for object detection tasks. We further verify our SNN on FPGA platform and the proposed model can achieve 800+FPS object detection with extremely low latency.
Zero-Shot and Few-Shot Video Question Answering with Multi-Modal Prompts
Authors: Deniz Engin, Yannis Avrithis
Subjects: Computer Vision and Pattern Recognition (cs.CV)
Abstract
Recent vision-language models are driven by large-scale pretrained models. However, adapting pretrained models on limited data presents challenges such as overfitting, catastrophic forgetting, and the cross-modal gap between vision and language. We introduce a parameter-efficient method to address these challenges, combining multimodal prompt learning and a transformer-based mapping network, while keeping the pretrained models frozen. Our experiments on several video question answering benchmarks demonstrate the superiority of our approach in terms of performance and parameter efficiency on both zero-shot and few-shot settings. Our code is available at https://engindeniz.github.io/vitis.
Towards Efficient and Trustworthy AI Through Hardware-Algorithm-Communication Co-Design
Authors: Bipin Rajendran, Osvaldo Simeone, Bashir M. Al-Hashimi
Subjects: Artificial Intelligence (cs.AI); Emerging Technologies (cs.ET); Information Theory (cs.IT)
Abstract
Artificial intelligence (AI) algorithms based on neural networks have been designed for decades with the goal of maximising some measure of accuracy. This has led to two undesired effects. First, model complexity has risen exponentially when measured in terms of computation and memory requirements. Second, state-of-the-art AI models are largely incapable of providing trustworthy measures of their uncertainty, possibly `hallucinating' their answers and discouraging their adoption for decision-making in sensitive applications. With the goal of realising efficient and trustworthy AI, in this paper we highlight research directions at the intersection of hardware and software design that integrate physical insights into computational substrates, neuroscientific principles concerning efficient information processing, information-theoretic results on optimal uncertainty quantification, and communication-theoretic guidelines for distributed processing. Overall, the paper advocates for novel design methodologies that target not only accuracy but also uncertainty quantification, while leveraging emerging computing hardware architectures that move beyond the traditional von Neumann digital computing paradigm to embrace in-memory, neuromorphic, and quantum computing technologies. An important overarching principle of the proposed approach is to view the stochasticity inherent in the computational substrate and in the communication channels between processors as a resource to be leveraged for the purpose of representing and processing classical and quantum uncertainty.
Scalable Multi-Robot Collaboration with Large Language Models: Centralized or Decentralized Systems?
Authors: Yongchao Chen, Jacob Arkin, Yang Zhang, Nicholas Roy, Chuchu Fan
Abstract
A flurry of recent work has demonstrated that pre-trained large language models (LLMs) can be effective task planners for a variety of single-robot tasks. The planning performance of LLMs is significantly improved via prompting techniques, such as in-context learning or re-prompting with state feedback, placing new importance on the token budget for the context window. An under-explored but natural next direction is to investigate LLMs as multi-robot task planners. However, long-horizon, heterogeneous multi-robot planning introduces new challenges of coordination while also pushing up against the limits of context window length. It is therefore critical to find token-efficient LLM planning frameworks that are also able to reason about the complexities of multi-robot coordination. In this work, we compare the task success rate and token efficiency of four multi-agent communication frameworks (centralized, decentralized, and two hybrid) as applied to four coordination-dependent multi-agent 2D task scenarios for increasing numbers of agents. We find that a hybrid framework achieves better task success rates across all four tasks and scales better to more agents. We further demonstrate the hybrid frameworks in 3D simulations where the vision-to-text problem and dynamical errors are considered. See our project website https://yongchao98.github.io/MIT-REALM-Multi-Robot/ for prompts, videos, and code.
Accelerating Motion Planning via Optimal Transport
Authors: An T. Le, Georgia Chalvatzaki, Armin Biess, Jan Peters
Subjects: Robotics (cs.RO); Optimization and Control (math.OC)
Abstract
Motion planning is still an open problem for many disciplines, e.g., robotics, autonomous driving, due to issues like high planning times that hinder real-time, efficient decision-making. A class of methods striving to provide smooth solutions is gradient-based trajectory optimization. However, those methods might suffer from bad local minima, while for many settings, they may be inapplicable due to the absence of easy access to objectives-gradients. In response to these issues, we introduce Motion Planning via Optimal Transport (MPOT) - a gradient-free method that optimizes a batch of smooth trajectories over highly nonlinear costs, even for high-dimensional tasks, while imposing smoothness through a Gaussian Process dynamics prior via planning-as-inference perspective. To facilitate batch trajectory optimization, we introduce an original zero-order and highly-parallelizable update rule -- the Sinkhorn Step, which uses the regular polytope family for its search directions; each regular polytope, centered on trajectory waypoints, serves as a local neighborhood, effectively acting as a trust region, where the Sinkhorn Step "transports" local waypoints toward low-cost regions. We theoretically show that Sinkhorn Step guides the optimizing parameters toward local minima regions on non-convex objective functions. We then show the efficiency of MPOT in a range of problems from low-dimensional point-mass navigation to high-dimensional whole-body robot motion planning, evincing its superiority compared with popular motion planners and paving the way for new applications of optimal transport in motion planning.
Loupe: Driving the Development of OS Compatibility Layers
Authors: Hugo Lefeuvre, Gaulthier Gain, Vlad-Andrei Bădoiu, Daniel Dinca, Vlad-Radu Schiller, Costin Raiciu, Felipe Huici, Pierre Olivier
Abstract
Supporting mainstream applications is fundamental for a new OS to have impact. It is generally achieved by developing a layer of compatibility allowing applications developed for a mainstream OS like Linux to run unmodified on the new OS. Building such a layer, as we show, results in large engineering inefficiencies due to the lack of efficient methods to precisely measure the OS features required by a set of applications. We propose Loupe, a novel method based on dynamic analysis that determines the OS features that need to be implemented in a prototype OS to bring support for a target set of applications and workloads. Loupe guides and boosts OS developers as they build compatibility layers, prioritizing which features to implement in order to quickly support many applications as early as possible. We apply our methodology to 100+ applications and several OSes currently under development, demonstrating high engineering effort savings vs. existing approaches: for example, for the 62 applications supported by the OSv kernel, we show that using Loupe, would have required implementing only 37 system calls vs. 92 for the non-systematic process followed by OSv developers. We study our measurements and extract novel key insights. Overall, we show that the burden of building compatibility layers is significantly less than what previous works suggest: in some cases, only as few as 20% of system calls reported by static analysis, and 50% of those reported by naive dynamic analysis need an implementation for an application to successfully run standard benchmarks.
Robust Blockwise Random Pivoting: Fast and Accurate Adaptive Interpolative Decomposition
Abstract
The interpolative decomposition (ID) aims to construct a low-rank approximation formed by a basis consisting of row/column skeletons in the original matrix and a corresponding interpolation matrix. This work explores fast and accurate ID algorithms from five essential perspectives for empirical performance: (a) skeleton complexity that measures the minimum possible ID rank for a given low-rank approximation error, (b) asymptotic complexity in FLOPs, (c) parallelizability of the computational bottleneck as matrix-matrix multiplications, (d) error-revealing property that enables automatic rank detection for given error tolerances without prior knowledge of target ranks, (e) ID-revealing property that ensures efficient construction of the optimal interpolation matrix after selecting the skeletons. While a broad spectrum of algorithms have been developed to optimize parts of the aforementioned perspectives, practical ID algorithms proficient in all perspectives remain absent. To fill in the gap, we introduce robust blockwise random pivoting (RBRP) that is parallelizable, error-revealing, and exact-ID-revealing, with comparable skeleton and asymptotic complexities to the best existing ID algorithms in practice. Through extensive numerical experiments on various synthetic and natural datasets, we empirically demonstrate the appealing performance of RBRP from the five perspectives above, as well as the robustness of RBRP to adversarial inputs.
GNNHLS: Evaluating Graph Neural Network Inference via High-Level Synthesis
Authors: Chenfeng Zhao, Zehao Dong, Yixin Chen, Xuan Zhang, Roger D. Chamberlain
Abstract
With the ever-growing popularity of Graph Neural Networks (GNNs), efficient GNN inference is gaining tremendous attention. Field-Programming Gate Arrays (FPGAs) are a promising execution platform due to their fine-grained parallelism, low-power consumption, reconfigurability, and concurrent execution. Even better, High-Level Synthesis (HLS) tools bridge the gap between the non-trivial FPGA development efforts and rapid emergence of new GNN models. In this paper, we propose GNNHLS, an open-source framework to comprehensively evaluate GNN inference acceleration on FPGAs via HLS, containing a software stack for data generation and baseline deployment, and FPGA implementations of 6 well-tuned GNN HLS kernels. We evaluate GNNHLS on 4 graph datasets with distinct topologies and scales. The results show that GNNHLS achieves up to 50.8x speedup and 423x energy reduction relative to the CPU baselines. Compared with the GPU baselines, GNNHLS achieves up to 5.16x speedup and 74.5x energy reduction.
Symbolic Imitation Learning: From Black-Box to Explainable Driving Policies
Abstract
Current methods of imitation learning (IL), primarily based on deep neural networks, offer efficient means for obtaining driving policies from real-world data but suffer from significant limitations in interpretability and generalizability. These shortcomings are particularly concerning in safety-critical applications like autonomous driving. In this paper, we address these limitations by introducing Symbolic Imitation Learning (SIL), a groundbreaking method that employs Inductive Logic Programming (ILP) to learn driving policies which are transparent, explainable and generalisable from available datasets. Utilizing the real-world highD dataset, we subject our method to a rigorous comparative analysis against prevailing neural-network-based IL methods. Our results demonstrate that SIL not only enhances the interpretability of driving policies but also significantly improves their applicability across varied driving situations. Hence, this work offers a novel pathway to more reliable and safer autonomous driving systems, underscoring the potential of integrating ILP into the domain of IL.
Bridging the complexity gap in Tbps-achieving THz-band baseband processing
Authors: Hadi Sarieddeen, Hakim Jemaa, Simon Tarboush, Christoph Studer, Mohamed-Slim Alouini, Tareq Y. Al-Naffouri
Subjects: Information Theory (cs.IT); Signal Processing (eess.SP)
Abstract
Recent advances in electronic and photonic technologies have allowed efficient signal generation and transmission at terahertz (THz) frequencies. However, as the gap in THz-operating devices narrows, the demand for terabit-per-second (Tbps)-achieving circuits is increasing. Translating the available hundreds of gigahertz (GHz) of bandwidth into a Tbps data rate requires processing thousands of information bits per clock cycle at state-of-the-art clock frequencies of digital baseband processing circuitry of a few GHz. This paper addresses these constraints and emphasizes the importance of parallelization in signal processing, particularly for channel code decoding. By leveraging structured sub-spaces of THz channels, we propose mapping bits to transmission resources using shorter code words, extending parallelizability across all baseband processing blocks. THz channels exhibit quasi-deterministic frequency, time, and space structures that enable efficient parallel bit mapping at the source and provide pseudo-soft bit reliability information for efficient detection and decoding at the receiver.
Channel Estimation for Reconfigurable Intelligent Surface-Aided Multiuser Communication Systems Exploiting Statistical CSI of Correlated RIS-User Channels
Authors: Haochen Li, Zhiwen Pan, Bin Wang, Nan Liu, Xiaohu You
Subjects: Information Theory (cs.IT); Signal Processing (eess.SP)
Abstract
Reconfigurable intelligent surface (RIS) is a promising candidate technology for the upcoming Sixth Generation (6G) communication system for its ability to manipulate the wireless communication environment by controlling the coefficients of reflection elements (REs). However, since the RIS usually consists of a large number of passive REs, the pilot overhead for channel estimation in the RIS-aided system is prohibitively high. In this paper, the channel estimation problem for a RIS-aided multi-user multiple-input-single-output (MISO) communication system with clustered users is investigated. First, to describe the correlated feature for RIS-user channels, a beam domain channel model is developed for RIS-user channels. Then, a pilot reuse strategy is put forward to reduce the pilot overhead and decompose the channel estimation problem into several subproblems. Finally, by leveraging the correlated nature of RIS-user channels, an eigenspace projection (EP) algorithm is proposed to solve each subproblem respectively. Simulation results show that the proposed EP channel estimation scheme can achieve accurate channel estimation with lower pilot overhead than existing schemes.
Effective Long-Context Scaling of Foundation Models
Authors: Wenhan Xiong, Jingyu Liu, Igor Molybog, Hejia Zhang, Prajjwal Bhargava, Rui Hou, Louis Martin, Rashi Rungta, Karthik Abinav Sankararaman, Barlas Oguz, Madian Khabsa, Han Fang, Yashar Mehdad, Sharan Narang, Kshitiz Malik, Angela Fan, Shruti Bhosale, Sergey Edunov, Mike Lewis, Sinong Wang, Hao Ma
Abstract
We present a series of long-context LLMs that support effective context windows of up to 32,768 tokens. Our model series are built through continual pretraining from Llama 2 with longer training sequences and on a dataset where long texts are upsampled. We perform extensive evaluation on language modeling, synthetic context probing tasks, and a wide range of research benchmarks. On research benchmarks, our models achieve consistent improvements on most regular tasks and significant improvements on long-context tasks over Llama 2. Notably, with a cost-effective instruction tuning procedure that does not require human-annotated long instruction data, the 70B variant can already surpass gpt-3.5-turbo-16k's overall performance on a suite of long-context tasks. Alongside these results, we provide an in-depth analysis on the individual components of our method. We delve into Llama's position encodings and discuss its limitation in modeling long dependencies. We also examine the impact of various design choices in the pretraining process, including the data mix and the training curriculum of sequence lengths -- our ablation experiments suggest that having abundant long texts in the pretrain dataset is not the key to achieving strong performance, and we empirically verify that long context continual pretraining is more efficient and similarly effective compared to pretraining from scratch with long sequences.
A Multiscale Finite Element Method for an Elliptic Distributed Optimal Control Problem with Rough Coefficients and Control Constraints
Authors: Susanne C. Brenner, Jose C. Garay, Li-yeng Sung
Abstract
We construct and analyze a multiscale finite element method for an elliptic distributed optimal control problem with pointwise control constraints, where the state equation has rough coefficients. We show that the performance of the multiscale finite element method is similar to the performance of standard finite element methods for smooth problems and present corroborating numerical results.
Label Augmentation Method for Medical Landmark Detection in Hip Radiograph Images
Authors: Yehyun Suh, Peter Chan, J.Ryan Martin, Daniel Moyer
Abstract
This work reports the empirical performance of an automated medical landmark detection method for predict clinical markers in hip radiograph images. Notably, the detection method was trained using a label-only augmentation scheme; our results indicate that this form of augmentation outperforms traditional data augmentation and produces highly sample efficient estimators. We train a generic U-Net-based architecture under a curriculum consisting of two phases: initially relaxing the landmarking task by enlarging the label points to regions, then gradually eroding these label regions back to the base task. We measure the benefits of this approach on six datasets of radiographs with gold-standard expert annotations.
A Modular Bio-inspired Robotic Hand with High Sensitivity
Authors: Chao Liu, Andrea Moncada, Hanna Matusik, Deniz Irem Erus, Daniela Rus
Abstract
While parallel grippers and multi-fingered robotic hands are well developed and commonly used in structured settings, it remains a challenge in robotics to design a highly articulated robotic hand that can be comparable to human hands to handle various daily manipulation and grasping tasks. Dexterity usually requires more actuators but also leads to a more sophisticated mechanism design and is more expensive to fabricate and maintain. Soft materials are able to provide compliance and safety when interacting with the physical world but are hard to model. This work presents a hybrid bio-inspired robotic hand that combines soft matters and rigid elements. Sensing is integrated into the rigid bodies resulting in a simple way for pose estimation with high sensitivity. The proposed hand is in a modular structure allowing for rapid fabrication and programming. The fabrication process is carefully designed so that a full hand can be made with low-cost materials and assembled in an efficient manner. We demonstrate the dexterity of the hand by successfully performing human grasp types.
A priori and a posteriori error analysis for a VEM discretization of the convection-diffusion eigenvalue problem
Authors: Danilo Amigo, Felipe Lepe, Gonzalo Rivera
Abstract
In this paper we propose and analyze a virtual element method for the two dimensional non-symmetric diffusion-convection eigenvalue problem in order to derive a priori and a posteriori error estimates. Under the classic assumptions of the meshes, and with the aid of the classic theory of compact operators, we prove error estimates for the eigenvalues and eigenfunctions. Also, we develop an a posteriori error estimator which, in one hand, results to be reliable and on the other, with standard bubble functions arguments, also results to be efficient. We test our method on domains where the complex eigenfunctions are not sufficiently regular, in order to assess the performance of the estimator that we compare with the uniform refinement given by the a priori analysis
Differentiable Robot Neural Distance Function for Adaptive Grasp Synthesis on a Unified Robotic Arm-Hand System
Abstract
Grasping is a fundamental skill for robots to interact with their environment. While grasp execution requires coordinated movement of the hand and arm to achieve a collision-free and secure grip, many grasp synthesis studies address arm and hand motion planning independently, leading to potentially unreachable grasps in practical settings. The challenge of determining integrated arm-hand configurations arises from its computational complexity and high-dimensional nature. We address this challenge by presenting a novel differentiable robot neural distance function. Our approach excels in capturing intricate geometry across various joint configurations while preserving differentiability. This innovative representation proves instrumental in efficiently addressing downstream tasks with stringent contact constraints. Leveraging this, we introduce an adaptive grasp synthesis framework that exploits the full potential of the unified arm-hand system for diverse grasping tasks. Our neural joint space distance function achieves an 84.7% error reduction compared to baseline methods. We validated our approaches on a unified robotic arm-hand system that consists of a 7-DoF robot arm and a 16-DoF multi-fingered robotic hand. Results demonstrate that our approach empowers this high-DoF system to generate and execute various arm-hand grasp configurations that adapt to the size of the target objects while ensuring whole-body movements to be collision-free.
Discovering Utility-driven Interval Rules
Authors: Chunkai Zhang, Maohua Lyu, Huaijin Hao, Wensheng Gan, Philip S. Yu
Abstract
For artificial intelligence, high-utility sequential rule mining (HUSRM) is a knowledge discovery method that can reveal the associations between events in the sequences. Recently, abundant methods have been proposed to discover high-utility sequence rules. However, the existing methods are all related to point-based sequences. Interval events that persist for some time are common. Traditional interval-event sequence knowledge discovery tasks mainly focus on pattern discovery, but patterns cannot reveal the correlation between interval events well. Moreover, the existing HUSRM algorithms cannot be directly applied to interval-event sequences since the relation in interval-event sequences is much more intricate than those in point-based sequences. In this work, we propose a utility-driven interval rule mining (UIRMiner) algorithm that can extract all utility-driven interval rules (UIRs) from the interval-event sequence database to solve the problem. In UIRMiner, we first introduce a numeric encoding relation representation, which can save much time on relation computation and storage on relation representation. Furthermore, to shrink the search space, we also propose a complement pruning strategy, which incorporates the utility upper bound with the relation. Finally, plentiful experiments implemented on both real-world and synthetic datasets verify that UIRMiner is an effective and efficient algorithm.
Learning Effective NeRFs and SDFs Representations with 3D Generative Adversarial Networks for 3D Object Generation: Technical Report for ICCV 2023 OmniObject3D Challenge
Authors: Zheyuan Yang, Yibo Liu, Guile Wu, Tongtong Cao, Yuan Ren, Yang Liu, Bingbing Liu
Subjects: Computer Vision and Pattern Recognition (cs.CV)
Abstract
In this technical report, we present a solution for 3D object generation of ICCV 2023 OmniObject3D Challenge. In recent years, 3D object generation has made great process and achieved promising results, but it remains a challenging task due to the difficulty of generating complex, textured and high-fidelity results. To resolve this problem, we study learning effective NeRFs and SDFs representations with 3D Generative Adversarial Networks (GANs) for 3D object generation. Specifically, inspired by recent works, we use the efficient geometry-aware 3D GANs as the backbone incorporating with label embedding and color mapping, which enables to train the model on different taxonomies simultaneously. Then, through a decoder, we aggregate the resulting features to generate Neural Radiance Fields (NeRFs) based representations for rendering high-fidelity synthetic images. Meanwhile, we optimize Signed Distance Functions (SDFs) to effectively represent objects with 3D meshes. Besides, we observe that this model can be effectively trained with only a few images of each object from a variety of classes, instead of using a great number of images per object or training one model per class. With this pipeline, we can optimize an effective model for 3D object generation. This solution is one of the final top-3-place solutions in the ICCV 2023 OmniObject3D Challenge.
E2Net: Resource-Efficient Continual Learning with Elastic Expansion Network
Authors: RuiQi Liu, Boyu Diao, Libo Huang, Zhulin An, Yongjun Xu
Abstract
Continual Learning methods are designed to learn new tasks without erasing previous knowledge. However, Continual Learning often requires massive computational power and storage capacity for satisfactory performance. In this paper, we propose a resource-efficient continual learning method called the Elastic Expansion Network (E2Net). Leveraging core subnet distillation and precise replay sample selection, E2Net achieves superior average accuracy and diminished forgetting within the same computational and storage constraints, all while minimizing processing time. In E2Net, we propose Representative Network Distillation to identify the representative core subnet by assessing parameter quantity and output similarity with the working network, distilling analogous subnets within the working network to mitigate reliance on rehearsal buffers and facilitating knowledge transfer across previous tasks. To enhance storage resource utilization, we then propose Subnet Constraint Experience Replay to optimize rehearsal efficiency through a sample storage strategy based on the structures of representative networks. Extensive experiments conducted predominantly on cloud environments with diverse datasets and also spanning the edge environment demonstrate that E2Net consistently outperforms state-of-the-art methods. In addition, our method outperforms competitors in terms of both storage and computational requirements.
ModuLoRA: Finetuning 3-Bit LLMs on Consumer GPUs by Integrating with Modular Quantizers
Authors: Junjie Yin, Jiahao Dong, Yingheng Wang, Christopher De Sa, Volodymyr Kuleshov
Abstract
We propose a memory-efficient finetuning algorithm for large language models (LLMs) that supports finetuning LLMs with 65B parameters in 3-bit or 4-bit precision on as little as one 48GB GPU. Our method, modular low-rank adaptation (ModuLoRA), integrates any user-specified weight quantizer with finetuning via low-rank adapters (LoRAs). Our approach relies on a simple quantization-agnostic backward pass that adaptively materializes low-precision LLM weights from a custom black-box quantization module. This approach enables finetuning 3-bit LLMs for the first time--leveraging state-of-the-art 3-bit OPTQ quantization often outperforms finetuning that relies on less sophisticated 4-bit and 8-bit methods. In our experiments, ModuLoRA attains competitive performance on text classification, natural language infernece, and instruction following tasks using significantly less memory than existing approaches, and we also surpass the state-of-the-art ROUGE score on a popular summarization task. We release ModuLoRA together with a series of low-precision models--including the first family of 3-bit instruction following Alpaca LLMs--as part of LLMTOOLS, a user-friendly library for quantizing, running, and finetuning LLMs on consumer GPUs.
UVL: A Unified Framework for Video Tampering Localization
Authors: Pengfei Pei, Xianfeng Zhao, Jinchuan Li, Yun Cao
Subjects: Computer Vision and Pattern Recognition (cs.CV); Cryptography and Security (cs.CR)
Abstract
With the development of deep learning technology, various forgery methods emerge endlessly. Meanwhile, methods to detect these fake videos have also achieved excellent performance on some datasets. However, these methods suffer from poor generalization to unknown videos and are inefficient for new forgery methods. To address this challenging problem, we propose UVL, a novel unified video tampering localization framework for synthesizing forgeries. Specifically, UVL extracts common features of synthetic forgeries: boundary artifacts of synthetic edges, unnatural distribution of generated pixels, and noncorrelation between the forgery region and the original. These features are widely present in different types of synthetic forgeries and help improve generalization for detecting unknown videos. Extensive experiments on three types of synthetic forgery: video inpainting, video splicing and DeepFake show that the proposed UVL achieves state-of-the-art performance on various benchmarks and outperforms existing methods by a large margin on cross-dataset.
Generative Semi-supervised Learning with Meta-Optimized Synthetic Samples
Abstract
Semi-supervised learning (SSL) is a promising approach for training deep classification models using labeled and unlabeled datasets. However, existing SSL methods rely on a large unlabeled dataset, which may not always be available in many real-world applications due to legal constraints (e.g., GDPR). In this paper, we investigate the research question: Can we train SSL models without real unlabeled datasets? Instead of using real unlabeled datasets, we propose an SSL method using synthetic datasets generated from generative foundation models trained on datasets containing millions of samples in diverse domains (e.g., ImageNet). Our main concepts are identifying synthetic samples that emulate unlabeled samples from generative foundation models and training classifiers using these synthetic samples. To achieve this, our method is formulated as an alternating optimization problem: (i) meta-learning of generative foundation models and (ii) SSL of classifiers using real labeled and synthetic unlabeled samples. For (i), we propose a meta-learning objective that optimizes latent variables to generate samples that resemble real labeled samples and minimize the validation loss. For (ii), we propose a simple unsupervised loss function that regularizes the feature extractors of classifiers to maximize the performance improvement obtained from synthetic samples. We confirm that our method outperforms baselines using generative foundation models on SSL. We also demonstrate that our methods outperform SSL using real unlabeled datasets in scenarios with extremely small amounts of labeled datasets. This suggests that synthetic samples have the potential to provide improvement gains more efficiently than real unlabeled data.
The Trickle-down Impact of Reward (In-)consistency on RLHF
Abstract
Standard practice within Reinforcement Learning from Human Feedback (RLHF) involves optimizing against a Reward Model (RM), which itself is trained to reflect human preferences for desirable generations. A notable subject that is understudied is the (in-)consistency of RMs -- whether they can recognize the semantic changes to different prompts and appropriately adapt their reward assignments -- and their impact on the downstream RLHF model. In this paper, we visit a series of research questions relevant to RM inconsistency: (1) How can we measure the consistency of reward models? (2) How consistent are the existing RMs and how can we improve them? (3) In what ways does reward inconsistency influence the chatbots resulting from the RLHF model training? We propose Contrast Instructions -- a benchmarking strategy for the consistency of RM. Each example in Contrast Instructions features a pair of lexically similar instructions with different ground truth responses. A consistent RM is expected to rank the corresponding instruction and response higher than other combinations. We observe that current RMs trained with the standard ranking objective fail miserably on Contrast Instructions compared to average humans. To show that RM consistency can be improved efficiently without using extra training budget, we propose two techniques ConvexDA and RewardFusion, which enhance reward consistency through extrapolation during the RM training and inference stage, respectively. We show that RLHF models trained with a more consistent RM yield more useful responses, suggesting that reward inconsistency exhibits a trickle-down effect on the downstream RLHF process.
Doppler Time-of-Flight Rendering
Authors: Juhyeon Kim, Wojciech Jarosz, Ioannis Gkioulekas, Adithya Pediredla
Abstract
We introduce Doppler time-of-flight (D-ToF) rendering, an extension of ToF rendering for dynamic scenes, with applications in simulating D-ToF cameras. D-ToF cameras use high-frequency modulation of illumination and exposure, and measure the Doppler frequency shift to compute the radial velocity of dynamic objects. The time-varying scene geometry and high-frequency modulation functions used in such cameras make it challenging to accurately and efficiently simulate their measurements with existing ToF rendering algorithms. We overcome these challenges in a twofold manner: To achieve accuracy, we derive path integral expressions for D-ToF measurements under global illumination and form unbiased Monte Carlo estimates of these integrals. To achieve efficiency, we develop a tailored time-path sampling technique that combines antithetic time sampling with correlated path sampling. We show experimentally that our sampling technique achieves up to two orders of magnitude lower variance compared to naive time-path sampling. We provide an open-source simulator that serves as a digital twin for D-ToF imaging systems, allowing imaging researchers, for the first time, to investigate the impact of modulation functions, material properties, and global illumination on D-ToF imaging performance.
Learning to Terminate in Object Navigation
Authors: Yuhang Song, Anh Nguyen, Chun-Yi Lee
Subjects: Robotics (cs.RO); Computer Vision and Pattern Recognition (cs.CV)
Abstract
This paper tackles the critical challenge of object navigation in autonomous navigation systems, particularly focusing on the problem of target approach and episode termination in environments with long optimal episode length in Deep Reinforcement Learning (DRL) based methods. While effective in environment exploration and object localization, conventional DRL methods often struggle with optimal path planning and termination recognition due to a lack of depth information. To overcome these limitations, we propose a novel approach, namely the Depth-Inference Termination Agent (DITA), which incorporates a supervised model called the Judge Model to implicitly infer object-wise depth and decide termination jointly with reinforcement learning. We train our judge model along with reinforcement learning in parallel and supervise the former efficiently by reward signal. Our evaluation shows the method is demonstrating superior performance, we achieve a 9.3% gain on success rate than our baseline method across all room types and gain 51.2% improvements on long episodes environment while maintaining slightly better Success Weighted by Path Length (SPL). Code and resources, visualization are available at: https://github.com/HuskyKingdom/DITA_acml2023
Max-Sliced Mutual Information
Authors: Dor Tsur, Ziv Goldfeld, Kristjan Greenewald
Subjects: Machine Learning (cs.LG); Information Theory (cs.IT)
Abstract
Quantifying the dependence between high-dimensional random variables is central to statistical learning and inference. Two classical methods are canonical correlation analysis (CCA), which identifies maximally correlated projected versions of the original variables, and Shannon's mutual information, which is a universal dependence measure that also captures high-order dependencies. However, CCA only accounts for linear dependence, which may be insufficient for certain applications, while mutual information is often infeasible to compute/estimate in high dimensions. This work proposes a middle ground in the form of a scalable information-theoretic generalization of CCA, termed max-sliced mutual information (mSMI). mSMI equals the maximal mutual information between low-dimensional projections of the high-dimensional variables, which reduces back to CCA in the Gaussian case. It enjoys the best of both worlds: capturing intricate dependencies in the data while being amenable to fast computation and scalable estimation from samples. We show that mSMI retains favorable structural properties of Shannon's mutual information, like variational forms and identification of independence. We then study statistical estimation of mSMI, propose an efficiently computable neural estimator, and couple it with formal non-asymptotic error bounds. We present experiments that demonstrate the utility of mSMI for several tasks, encompassing independence testing, multi-view representation learning, algorithmic fairness, and generative modeling. We observe that mSMI consistently outperforms competing methods with little-to-no computational overhead.
Hybrid Digital-Wave Domain Channel Estimator for Stacked Intelligent Metasurface Enabled Multi-User MISO Systems
Authors: Qurrat-Ul-Ain Nadeem, Jiancheng An, Anas Chaaban
Subjects: Information Theory (cs.IT); Signal Processing (eess.SP)
Abstract
Stacked intelligent metasurface (SIM) is an emerging programmable metasurface architecture that can implement signal processing directly in the electromagnetic wave domain, thereby enabling efficient implementation of ultra-massive multiple-input multiple-output (MIMO) transceivers with a limited number of radio frequency (RF) chains. Channel estimation (CE) is challenging for SIM-enabled communication systems due to the multi-layer architecture of SIM, and because we need to estimate large dimensional channels between the SIM and users with a limited number of RF chains. To efficiently solve this problem, we develop a novel hybrid digital-wave domain channel estimator, in which the received training symbols are first processed in the wave domain within the SIM layers, and then processed in the digital domain. The wave domain channel estimator, parametrized by the phase shifts applied by the meta-atoms in all layers, is optimized to minimize the mean squared error (MSE) using a gradient descent algorithm, within which the digital part is optimally updated. For an SIM-enabled multi-user system equipped with 4 RF chains and a 6-layer SIM with 64 meta-atoms each, the proposed estimator yields an MSE that is very close to that achieved by fully digital CE in a massive MIMO system employing 64 RF chains. This high CE accuracy is achieved at the cost of a training overhead that can be reduced by exploiting the potential low rank of channel correlation matrices.
DiffGAN-F2S: Symmetric and Efficient Denoising Diffusion GANs for Structural Connectivity Prediction from Brain fMRI
Authors: Qiankun Zuo, Ruiheng Li, Yi Di, Hao Tian, Changhong Jing, Xuhang Chen, Shuqiang Wang
Subjects: Computer Vision and Pattern Recognition (cs.CV); Image and Video Processing (eess.IV)
Abstract
Mapping from functional connectivity (FC) to structural connectivity (SC) can facilitate multimodal brain network fusion and discover potential biomarkers for clinical implications. However, it is challenging to directly bridge the reliable non-linear mapping relations between SC and functional magnetic resonance imaging (fMRI). In this paper, a novel diffusision generative adversarial network-based fMRI-to-SC (DiffGAN-F2S) model is proposed to predict SC from brain fMRI in an end-to-end manner. To be specific, the proposed DiffGAN-F2S leverages denoising diffusion probabilistic models (DDPMs) and adversarial learning to efficiently generate high-fidelity SC through a few steps from fMRI. By designing the dual-channel multi-head spatial attention (DMSA) and graph convolutional modules, the symmetric graph generator first captures global relations among direct and indirect connected brain regions, then models the local brain region interactions. It can uncover the complex mapping relations between fMRI and structural connectivity. Furthermore, the spatially connected consistency loss is devised to constrain the generator to preserve global-local topological information for accurate intrinsic SC prediction. Testing on the public Alzheimer's Disease Neuroimaging Initiative (ADNI) dataset, the proposed model can effectively generate empirical SC-preserved connectivity from four-dimensional imaging data and shows superior performance in SC prediction compared with other related models. Furthermore, the proposed model can identify the vast majority of important brain regions and connections derived from the empirical method, providing an alternative way to fuse multimodal brain networks and analyze clinical disease.
Parameter-Saving Adversarial Training: Reinforcing Multi-Perturbation Robustness via Hypernetworks
Abstract
Adversarial training serves as one of the most popular and effective methods to defend against adversarial perturbations. However, most defense mechanisms only consider a single type of perturbation while various attack methods might be adopted to perform stronger adversarial attacks against the deployed model in real-world scenarios, e.g., $\ell2$ or $\ell\infty$. Defending against various attacks can be a challenging problem since multi-perturbation adversarial training and its variants only achieve suboptimal robustness trade-offs, due to the theoretical limit to multi-perturbation robustness for a single model. Besides, it is impractical to deploy large models in some storage-efficient scenarios. To settle down these drawbacks, in this paper we propose a novel multi-perturbation adversarial training framework, parameter-saving adversarial training (PSAT), to reinforce multi-perturbation robustness with an advantageous side effect of saving parameters, which leverages hypernetworks to train specialized models against a single perturbation and aggregate these specialized models to defend against multiple perturbations. Eventually, we extensively evaluate and compare our proposed method with state-of-the-art single/multi-perturbation robust methods against various latest attack methods on different datasets, showing the robustness superiority and parameter efficiency of our proposed method, e.g., for the CIFAR-10 dataset with ResNet-50 as the backbone, PSAT saves approximately 80\% of parameters with achieving the state-of-the-art robustness trade-off accuracy.
Controllable Text Generation with Residual Memory Transformer
Authors: Hanqing Zhang, Sun Si, Haiming Wu, Dawei Song
Abstract
Large-scale Causal Language Models (CLMs), e.g., GPT3 and ChatGPT, have brought great success in text generation. However, it is still an open challenge to control the generation process of CLM while balancing flexibility, control granularity, and generation efficiency. In this paper, we provide a new alternative for controllable text generation (CTG), by designing a non-intrusive, lightweight control plugin to accompany the generation of CLM at arbitrary time steps. The proposed control plugin, namely Residual Memory Transformer (RMT), has an encoder-decoder setup, which can accept any types of control conditions and cooperate with CLM through a residual learning paradigm, to achieve a more flexible, general, and efficient CTG. Extensive experiments are carried out on various control tasks, in the form of both automatic and human evaluations. The results show the superiority of RMT over a range of state-of-the-art approaches, proving the effectiveness and versatility of our approach.
Beyond Reverse KL: Generalizing Direct Preference Optimization with Diverse Divergence Constraints
Abstract
The increasing capabilities of large language models (LLMs) raise opportunities for artificial general intelligence but concurrently amplify safety concerns, such as potential misuse of AI systems, necessitating effective AI alignment. Reinforcement Learning from Human Feedback (RLHF) has emerged as a promising pathway towards AI alignment but brings forth challenges due to its complexity and dependence on a separate reward model. Direct Preference Optimization (DPO) has been proposed as an alternative, and it remains equivalent to RLHF under the reverse KL regularization constraint. This paper presents $f$-DPO, a generalized approach to DPO by incorporating diverse divergence constraints. We show that under certain $f$-divergences, including Jensen-Shannon divergence, forward KL divergences and $\alpha$-divergences, the complex relationship between the reward and optimal policy can also be simplified by addressing the Karush-Kuhn-Tucker conditions. This eliminates the need for estimating the normalizing constant in the Bradley-Terry model and enables a tractable mapping between the reward function and the optimal policy. Our approach optimizes LLMs to align with human preferences in a more efficient and supervised manner under a broad set of divergence constraints. Empirically, adopting these divergences ensures a balance between alignment performance and generation diversity. Importantly, $f$-DPO outperforms PPO-based methods in divergence efficiency, and divergence constraints directly influence expected calibration error (ECE).
Nondestructive chicken egg fertility detection using CNN-transfer learning algorithms
Authors: Shoffan Saifullah, Rafal Drezewski, Anton Yudhana, Andri Pranolo, Wilis Kaswijanti, Andiko Putro Suryotomo, Seno Aji Putra, Alin Khaliduzzaman, Anton Satria Prabuwono, Nathalie Japkowicz
Subjects: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI); Image and Video Processing (eess.IV)
Abstract
This study explored the application of CNN-Transfer Learning for nondestructive chicken egg fertility detection for precision poultry hatchery practices. Four models, VGG16, ResNet50, InceptionNet, and MobileNet, were trained and evaluated on a dataset (200 single egg images) using augmented images (rotation, flip, scale, translation, and reflection). Although the training results demonstrated that all models achieved high accuracy, indicating their ability to accurately learn and classify chicken eggs' fertility state, when evaluated on the testing set, variations in accuracy and performance were observed. InceptionNet exhibited the best overall performance, accurately classifying fertile and non-fertile eggs. It demonstrated excellent performance in both training and testing sets in all parameters of the evaluation metrics. In testing set, it achieved an accuracy of 0.98, a sensitivity of 1 for detecting fertile eggs, and a specificity of 0.96 for identifying non-fertile eggs. The higher performance is attributed to its unique architecture efficiently capturing features at different scales leading to improved accuracy and robustness. Further optimization and fine-tuning of the models might necessary to address the limitations in accurately detecting fertile and non-fertile eggs in case of other models. This study highlighted the potential of CNN-Transfer Learning for nondestructive fertility detection and emphasizes the need for further research to enhance the models' capabilities and ensure accurate classification.
Efficiency Separation between RL Methods: Model-Free, Model-Based and Goal-Conditioned
Abstract
We prove a fundamental limitation on the efficiency of a wide class of Reinforcement Learning (RL) algorithms. This limitation applies to model-free RL methods as well as a broad range of model-based methods, such as planning with tree search. Under an abstract definition of this class, we provide a family of RL problems for which these methods suffer a lower bound exponential in the horizon for their interactions with the environment to find an optimal behavior. However, there exists a method, not tailored to this specific family of problems, which can efficiently solve the problems in the family. In contrast, our limitation does not apply to several types of methods proposed in the literature, for instance, goal-conditioned methods or other algorithms that construct an inverse dynamics model.
Multi-scale Recurrent LSTM and Transformer Network for Depth Completion
Authors: Xiaogang Jia, Yusong Tan, Songlei Jian, Yonggang Che
Subjects: Computer Vision and Pattern Recognition (cs.CV)
Abstract
Lidar depth completion is a new and hot topic of depth estimation. In this task, it is the key and difficult point to fuse the features of color space and depth space. In this paper, we migrate the classic LSTM and Transformer modules from NLP to depth completion and redesign them appropriately. Specifically, we use Forget gate, Update gate, Output gate, and Skip gate to achieve the efficient fusion of color and depth features and perform loop optimization at multiple scales. Finally, we further fuse the deep features through the Transformer multi-head attention mechanism. Experimental results show that without repetitive network structure and post-processing steps, our method can achieve state-of-the-art performance by adding our modules to a simple encoder-decoder network structure. Our method ranks first on the current mainstream autonomous driving KITTI benchmark dataset. It can also be regarded as a backbone network for other methods, which likewise achieves state-of-the-art performance.
LagrangeBench: A Lagrangian Fluid Mechanics Benchmarking Suite
Authors: Artur P. Toshev, Gianluca Galletti, Fabian Fritz, Stefan Adami, Nikolaus A. Adams
Abstract
Machine learning has been successfully applied to grid-based PDE modeling in various scientific applications. However, learned PDE solvers based on Lagrangian particle discretizations, which are the preferred approach to problems with free surfaces or complex physics, remain largely unexplored. We present LagrangeBench, the first benchmarking suite for Lagrangian particle problems, focusing on temporal coarse-graining. In particular, our contribution is: (a) seven new fluid mechanics datasets (four in 2D and three in 3D) generated with the Smoothed Particle Hydrodynamics (SPH) method including the Taylor-Green vortex, lid-driven cavity, reverse Poiseuille flow, and dam break, each of which includes different physics like solid wall interactions or free surface, (b) efficient JAX-based API with various recent training strategies and neighbors search routine, and (c) JAX implementation of established Graph Neural Networks (GNNs) like GNS and SEGNN with baseline results. Finally, to measure the performance of learned surrogates we go beyond established position errors and introduce physical metrics like kinetic energy MSE and Sinkhorn distance for the particle distribution. Our codebase is available under the URL: https://github.com/tumaer/lagrangebench
Transformer-VQ: Linear-Time Transformers via Vector Quantization
Authors: Lucas D. Lingle
Subjects: Machine Learning (cs.LG); Computation and Language (cs.CL); Computer Vision and Pattern Recognition (cs.CV)
Abstract
We introduce Transformer-VQ, a decoder-only transformer computing softmax-based dense self-attention in linear time. Transformer-VQ's efficient attention is enabled by vector-quantized keys and a novel caching mechanism. In large-scale experiments, Transformer-VQ is shown highly competitive in quality, with strong results on Enwik8 (0.99 bpb), PG-19 (26.6 ppl), and ImageNet64 (3.16 bpb). Code: https://github.com/transformer-vq/transformer_vq
Nek5000/RS Performance on Advanced GPU Architectures
Authors: Misun Min, Yu-Hsiang Lan, Paul Fischer, Thilina Rathnayake, John Holmen
Subjects: Distributed, Parallel, and Cluster Computing (cs.DC); Performance (cs.PF)
Abstract
We demonstrate NekRS performance results on various advanced GPU architectures. NekRS is a GPU-accelerated version of Nek5000 that targets high performance on exascale platforms. It is being developed in DOE's Center of Efficient Exascale Discretizations, which is one of the co-design centers under the Exascale Computing Project. In this paper, we consider Frontier, Crusher, Spock, Polaris, Perlmutter, ThetaGPU, and Summit. Simulations are performed with 17x17 rod-bundle geometries from small modular reactor applications. We discuss strong-scaling performance and analysis.
Genetic Engineering Algorithm (GEA): An Efficient Metaheuristic Algorithm for Solving Combinatorial Optimization Problems
Authors: Majid Sohrabi, Amir M. Fathollahi-Fard, Vasilii A. Gromov
Subjects: Neural and Evolutionary Computing (cs.NE); Artificial Intelligence (cs.AI)
Abstract
Genetic Algorithms (GAs) are known for their efficiency in solving combinatorial optimization problems, thanks to their ability to explore diverse solution spaces, handle various representations, exploit parallelism, preserve good solutions, adapt to changing dynamics, handle combinatorial diversity, and provide heuristic search. However, limitations such as premature convergence, lack of problem-specific knowledge, and randomness of crossover and mutation operators make GAs generally inefficient in finding an optimal solution. To address these limitations, this paper proposes a new metaheuristic algorithm called the Genetic Engineering Algorithm (GEA) that draws inspiration from genetic engineering concepts. GEA redesigns the traditional GA while incorporating new search methods to isolate, purify, insert, and express new genes based on existing ones, leading to the emergence of desired traits and the production of specific chromosomes based on the selected genes. Comparative evaluations against state-of-the-art algorithms on benchmark instances demonstrate the superior performance of GEA, showcasing its potential as an innovative and efficient solution for combinatorial optimization problems.
Nonlinear MPC design for incrementally ISS systems with application to GRU networks
Authors: Fabio Bonassi, Alessio La Bella, Marcello Farina, Riccardo Scattolini
Subjects: Systems and Control (eess.SY); Machine Learning (cs.LG)
Abstract
This brief addresses the design of a Nonlinear Model Predictive Control (NMPC) strategy for exponentially incremental Input-to-State Stable (ISS) systems. In particular, a novel formulation is devised, which does not necessitate the onerous computation of terminal ingredients, but rather relies on the explicit definition of a minimum prediction horizon ensuring closed-loop stability. The designed methodology is particularly suited for the control of systems learned by Recurrent Neural Networks (RNNs), which are known for their enhanced modeling capabilities and for which the incremental ISS properties can be studied thanks to simple algebraic conditions. The approach is applied to Gated Recurrent Unit (GRU) networks, providing also a method for the design of a tailored state observer with convergence guarantees. The resulting control architecture is tested on a benchmark system, demonstrating its good control performances and efficient applicability.
A Small World of Bad Guys: Investigating the Behavior of Hacker Groups in Cyber-Attacks
Authors: Giampiero Giacomello, Antonio Iovanella, Luigi Martino
Abstract
This paper explores the behaviour of malicious hacker groups operating in cyberspace and how they organize themselves in structured networks. To better understand these groups, the paper uses Social Network Analysis (SNA) to analyse the interactions and relationships among several malicious hacker groups. The study uses a tested dataset as its primary source, providing an empirical analysis of the cooperative behaviours exhibited by these groups. The study found that malicious hacker groups tend to form close-knit networks where they consult, coordinate with, and assist each other in carrying out their attacks. The study also identified a "small world" phenomenon within the population of malicious actors, which suggests that these groups establish interconnected relationships to facilitate their malicious operations. The small world phenomenon indicates that the actor-groups are densely connected, but they also have a small number of connections to other groups, allowing for efficient communication and coordination of their activities.
Hadoop-Oriented SVM-LRU (H-SVM-LRU): An Intelligent Cache Replacement Algorithm to Improve MapReduce Performance
Authors: Rana Ghazali, Sahar Adabi, Ali Rezaee, Douglas G.Down, Ali Movaghar
Subjects: Distributed, Parallel, and Cluster Computing (cs.DC)
Abstract
Modern applications can generate a large amount of data from different sources with high velocity, a combination that is difficult to store and process via traditional tools. Hadoop is one framework that is used for the parallel processing of a large amount of data in a distributed environment, however, various challenges can lead to poor performance. Two particular issues that can limit performance are the high access time for I/O operations and the recomputation of intermediate data. The combination of these two issues can result in resource wastage. In recent years, there have been attempts to overcome these problems by using caching mechanisms. Due to cache space limitations, it is crucial to use this space efficiently and avoid cache pollution (the cache contains data that is not used in the future). We propose Hadoop-oriented SVM-LRU (HSVM- LRU) to improve Hadoop performance. For this purpose, we use an intelligent cache replacement algorithm, SVM-LRU, that combines the well-known LRU mechanism with a machine learning algorithm, SVM, to classify cached data into two groups based on their future usage. Experimental results show a significant decrease in execution time as a result of an increased cache hit ratio, leading to a positive impact on Hadoop performance.
Efficient Hardware Implementation of Constant Time Sampling for HQC
Authors: Maximilian Schöffel, Johannes Feldmann, Norbert Wehn
Abstract
HQC is one of the code-based finalists in the last round of the NIST post quantum cryptography standardization process. In this process, security and implementation efficiency are key metrics for the selection of the candidates. A critical compute kernel with respect to efficient hardware implementations and security in HQC is the sampling method used to derive random numbers. Due to its security criticality, recently an updated sampling algorithm was presented to increase its robustness against side-channel attacks. In this paper, we pursue a cross layer approach to optimize this new sampling algorithm to enable an efficient hardware implementation without comprising the original algorithmic security and side-channel attack robustness. We compare our cross layer based implementation to a direct hardware implementation of the original algorithm and to optimized implementations of the previous sampler version. All implementations are evaluated using the Xilinx Artix 7 FPGA. Our results show that our approach reduces the latency by a factor of 24 compared to the original algorithm and by a factor of 28 compared to the previously used sampler with significantly less resources.
Capturing Requirements for a Data Annotation Tool for Intensive Care: Experimental User-Centered Design Study
Authors: Marceli Wac, Raul Santos-Rodriguez, Chris McWilliams, Christopher Bourdeaux
Abstract
Intensive care units (ICUs) are complex and data-rich environments. Data routinely collected in the ICUs provides tremendous opportunities for machine learning, but their use comes with significant challenges. Complex problems may require additional input from humans which can be provided through a process of data annotation. Annotation is a complex, time-consuming process that requires domain expertise and technical proficiency. Existing data annotation tools fail to provide an effective solution to this problem. In this study, we investigated clinicians' approach to the annotation task. We focused on establishing the characteristics of the annotation process in the context of clinical data and identifying differences in the annotation workflow between different staff roles. The overall goal was to elicit requirements for a software tool that could facilitate an effective and time-efficient data annotation. We conducted an experiment involving clinicians from the ICUs annotating printed sheets of data. The participants were observed during the task and their actions were analysed in the context of Norman's Interaction Cycle to establish the requirements for the digital tool. The annotation process followed a constant loop of annotation and evaluation, during which participants incrementally analysed and annotated the data. No distinguishable differences were identified between how different staff roles annotate data. We observed preferences towards different methods for applying annotation which varied between different participants and admissions. We established 11 requirements for the digital data annotation tool for the healthcare setting. We conducted a manual data annotation activity to establish the requirements for a digital data annotation tool, characterised the clinicians' approach to annotation and elicited 11 key requirements for effective data annotation software.
SwiftSpatial: Spatial Joins on Modern Hardware
Authors: Wenqi Jiang, Martin Parvanov, Gustavo Alonso
Abstract
Spatial joins are among the most time-consuming queries in spatial data management systems. In this paper, we propose SwiftSpatial, a specialized accelerator architecture tailored for spatial joins. SwiftSpatial contains multiple high-performance join units with innovative hybrid parallelism, several efficient memory management units, and an integrated on-chip join scheduler. We prototype SwiftSpatial on an FPGA and incorporate the R-tree synchronous traversal algorithm as the control flow. Benchmarked against various CPU and GPU-based spatial data processing systems, SwiftSpatial demonstrates a latency reduction of up to 5.36x relative to the best-performing baseline, while requiring 6.16x less power. The remarkable performance and energy efficiency of SwiftSpatial lay a solid foundation for its future integration into spatial data management systems, both in data centers and at the edge.
HOI4ABOT: Human-Object Interaction Anticipation for Human Intention Reading Collaborative roBOTs
Authors: Esteve Valls Mascaro, Daniel Sliwowski, Dongheui Lee
Subjects: Computer Vision and Pattern Recognition (cs.CV); Robotics (cs.RO)
Abstract
Robots are becoming increasingly integrated into our lives, assisting us in various tasks. To ensure effective collaboration between humans and robots, it is essential that they understand our intentions and anticipate our actions. In this paper, we propose a Human-Object Interaction (HOI) anticipation framework for collaborative robots. We propose an efficient and robust transformer-based model to detect and anticipate HOIs from videos. This enhanced anticipation empowers robots to proactively assist humans, resulting in more efficient and intuitive collaborations. Our model outperforms state-of-the-art results in HOI detection and anticipation in VidHOI dataset with an increase of 1.76% and 1.04% in mAP respectively while being 15.4 times faster. We showcase the effectiveness of our approach through experimental results in a real robot, demonstrating that the robot's ability to anticipate HOIs is key for better Human-Robot Interaction. More information can be found on our project webpage: https://evm7.github.io/HOI4ABOT_page/
Audio-Visual Speaker Verification via Joint Cross-Attention
Authors: R. Gnana Praveen, Jahangir Alam
Subjects: Sound (cs.SD); Computer Vision and Pattern Recognition (cs.CV); Multimedia (cs.MM); Audio and Speech Processing (eess.AS)
Abstract
Speaker verification has been widely explored using speech signals, which has shown significant improvement using deep models. Recently, there has been a surge in exploring faces and voices as they can offer more complementary and comprehensive information than relying only on a single modality of speech signals. Though current methods in the literature on the fusion of faces and voices have shown improvement over that of individual face or voice modalities, the potential of audio-visual fusion is not fully explored for speaker verification. Most of the existing methods based on audio-visual fusion either rely on score-level fusion or simple feature concatenation. In this work, we have explored cross-modal joint attention to fully leverage the inter-modal complementary information and the intra-modal information for speaker verification. Specifically, we estimate the cross-attention weights based on the correlation between the joint feature presentation and that of the individual feature representations in order to effectively capture both intra-modal as well inter-modal relationships among the faces and voices. We have shown that efficiently leveraging the intra- and inter-modal relationships significantly improves the performance of audio-visual fusion for speaker verification. The performance of the proposed approach has been evaluated on the Voxceleb1 dataset. Results show that the proposed approach can significantly outperform the state-of-the-art methods of audio-visual fusion for speaker verification.
Compilation as a Defense: Enhancing DL Model Attack Robustness via Tensor Optimization
Authors: Stefan Trawicki, William Hackett, Lewis Birch, Neeraj Suri, Peter Garraghan
Subjects: Machine Learning (cs.LG); Cryptography and Security (cs.CR)
Abstract
Adversarial Machine Learning (AML) is a rapidly growing field of security research, with an often overlooked area being model attacks through side-channels. Previous works show such attacks to be serious threats, though little progress has been made on efficient remediation strategies that avoid costly model re-engineering. This work demonstrates a new defense against AML side-channel attacks using model compilation techniques, namely tensor optimization. We show relative model attack effectiveness decreases of up to 43% using tensor optimization, discuss the implications, and direction of future work.
Deep Learning Based Uplink Multi-User SIMO Beamforming Design
Authors: Cemil Vahapoglu, Timothy J. O'Shea, Tamoghna Roy, Sennur Ulukus
Subjects: Information Theory (cs.IT); Machine Learning (cs.LG); Signal Processing (eess.SP)
Abstract
The advancement of fifth generation (5G) wireless communication networks has created a greater demand for wireless resource management solutions that offer high data rates, extensive coverage, minimal latency and energy-efficient performance. Nonetheless, traditional approaches have shortcomings when it comes to computational complexity and their ability to adapt to dynamic conditions, creating a gap between theoretical analysis and the practical execution of algorithmic solutions for managing wireless resources. Deep learning-based techniques offer promising solutions for bridging this gap with their substantial representation capabilities. We propose a novel unsupervised deep learning framework, which is called NNBF, for the design of uplink receive multi-user single input multiple output (MU-SIMO) beamforming. The primary objective is to enhance the throughput by focusing on maximizing the sum-rate while also offering computationally efficient solution, in contrast to established conventional methods. We conduct experiments for several antenna configurations. Our experimental results demonstrate that NNBF exhibits superior performance compared to our baseline methods, namely, zero-forcing beamforming (ZFBF) and minimum mean square error (MMSE) equalizer. Additionally, NNBF is scalable to the number of single-antenna user equipments (UEs) while baseline methods have significant computational burden due to matrix pseudo-inverse operation.
Robust Offline Reinforcement Learning -- Certify the Confidence Interval
Abstract
Currently, reinforcement learning (RL), especially deep RL, has received more and more attention in the research area. However, the security of RL has been an obvious problem due to the attack manners becoming mature. In order to defend against such adversarial attacks, several practical approaches are developed, such as adversarial training, data filtering, etc. However, these methods are mostly based on empirical algorithms and experiments, without rigorous theoretical analysis of the robustness of the algorithms. In this paper, we develop an algorithm to certify the robustness of a given policy offline with random smoothing, which could be proven and conducted as efficiently as ones without random smoothing. Experiments on different environments confirm the correctness of our algorithm.
Sparse Submodular Function Minimization
Authors: Andrei Graur, Haotian Jiang, Aaron Sidford
Subjects: Data Structures and Algorithms (cs.DS); Optimization and Control (math.OC)
Abstract
In this paper we study the problem of minimizing a submodular function $f : 2^V \rightarrow \mathbb{R}$ that is guaranteed to have a $k$-sparse minimizer. We give a deterministic algorithm that computes an additive $\epsilon$-approximate minimizer of such $f$ in $\widetilde{O}(\mathsf{poly}(k) \log(|f|/\epsilon))$ parallel depth using a polynomial number of queries to an evaluation oracle of $f$, where $|f| = \max_{S \subseteq V} |f(S)|$. Further, we give a randomized algorithm that computes an exact minimizer of $f$ with high probability using $\widetilde{O}(|V| \cdot \mathsf{poly}(k))$ queries and polynomial time. When $k = \widetilde{O}(1)$, our algorithms use either nearly-constant parallel depth or a nearly-linear number of evaluation oracle queries. All previous algorithms for this problem either use $\Omega(|V|)$ parallel depth or $\Omega(|V|^2)$ queries. In contrast to state-of-the-art weakly-polynomial and strongly-polynomial time algorithms for SFM, our algorithms use first-order optimization methods, e.g., mirror descent and follow the regularized leader. We introduce what we call {\em sparse dual certificates}, which encode information on the structure of sparse minimizers, and both our parallel and sequential algorithms provide new algorithmic tools for allowing first-order optimization methods to efficiently compute them. Correspondingly, our algorithm does not invoke fast matrix multiplication or general linear system solvers and in this sense is more combinatorial than previous state-of-the-art methods.
ConceptGraphs: Open-Vocabulary 3D Scene Graphs for Perception and Planning
Authors: Qiao Gu, Alihusein Kuwajerwala, Sacha Morin, Krishna Murthy Jatavallabhula, Bipasha Sen, Aditya Agarwal, Corban Rivera, William Paul, Kirsty Ellis, Rama Chellappa, Chuang Gan, Celso Miguel de Melo, Joshua B. Tenenbaum, Antonio Torralba, Florian Shkurti, Liam Paull
Subjects: Robotics (cs.RO); Computer Vision and Pattern Recognition (cs.CV)
Abstract
For robots to perform a wide variety of tasks, they require a 3D representation of the world that is semantically rich, yet compact and efficient for task-driven perception and planning. Recent approaches have attempted to leverage features from large vision-language models to encode semantics in 3D representations. However, these approaches tend to produce maps with per-point feature vectors, which do not scale well in larger environments, nor do they contain semantic spatial relationships between entities in the environment, which are useful for downstream planning. In this work, we propose ConceptGraphs, an open-vocabulary graph-structured representation for 3D scenes. ConceptGraphs is built by leveraging 2D foundation models and fusing their output to 3D by multi-view association. The resulting representations generalize to novel semantic classes, without the need to collect large 3D datasets or finetune models. We demonstrate the utility of this representation through a number of downstream planning tasks that are specified through abstract (language) prompts and require complex reasoning over spatial and semantic concepts. (Project page: https://concept-graphs.github.io/ Explainer video: https://youtu.be/mRhNkQwRYnc )
DreamGaussian: Generative Gaussian Splatting for Efficient 3D Content Creation
Authors: Jiaxiang Tang, Jiawei Ren, Hang Zhou, Ziwei Liu, Gang Zeng
Subjects: Computer Vision and Pattern Recognition (cs.CV)
Abstract
Recent advances in 3D content creation mostly leverage optimization-based 3D generation via score distillation sampling (SDS). Though promising results have been exhibited, these methods often suffer from slow per-sample optimization, limiting their practical usage. In this paper, we propose DreamGaussian, a novel 3D content generation framework that achieves both efficiency and quality simultaneously. Our key insight is to design a generative 3D Gaussian Splatting model with companioned mesh extraction and texture refinement in UV space. In contrast to the occupancy pruning used in Neural Radiance Fields, we demonstrate that the progressive densification of 3D Gaussians converges significantly faster for 3D generative tasks. To further enhance the texture quality and facilitate downstream applications, we introduce an efficient algorithm to convert 3D Gaussians into textured meshes and apply a fine-tuning stage to refine the details. Extensive experiments demonstrate the superior efficiency and competitive generation quality of our proposed approach. Notably, DreamGaussian produces high-quality textured meshes in just 2 minutes from a single-view image, achieving approximately 10 times acceleration compared to existing methods.
Training a Large Video Model on a Single Machine in a Day
Authors: Yue Zhao, Philipp Krähenbühl
Subjects: Computer Vision and Pattern Recognition (cs.CV)
Abstract
Videos are big, complex to pre-process, and slow to train on. State-of-the-art large-scale video models are trained on clusters of 32 or more GPUs for several days. As a consequence, academia largely ceded the training of large video models to industry. In this paper, we show how to still train a state-of-the-art video model on a single machine with eight consumer-grade GPUs in a day. We identify three bottlenecks, IO, CPU, and GPU computation, and optimize each. The result is a highly efficient video training pipeline. For comparable architectures, our pipeline achieves higher accuracies with $\frac{1}{8}$ of the computation compared to prior work. Code is available at https://github.com/zhaoyue-zephyrus/AVION.
Keyword: faster
CLIP-Hand3D: Exploiting 3D Hand Pose Estimation via Context-Aware Prompting
Authors: Shaoxiang Guo, Qing Cai, Lin Qi, Junyu Dong
Subjects: Multimedia (cs.MM); Computer Vision and Pattern Recognition (cs.CV)
Abstract
Contrastive Language-Image Pre-training (CLIP) starts to emerge in many computer vision tasks and has achieved promising performance. However, it remains underexplored whether CLIP can be generalized to 3D hand pose estimation, as bridging text prompts with pose-aware features presents significant challenges due to the discrete nature of joint positions in 3D space. In this paper, we make one of the first attempts to propose a novel 3D hand pose estimator from monocular images, dubbed as CLIP-Hand3D, which successfully bridges the gap between text prompts and irregular detailed pose distribution. In particular, the distribution order of hand joints in various 3D space directions is derived from pose labels, forming corresponding text prompts that are subsequently encoded into text representations. Simultaneously, 21 hand joints in the 3D space are retrieved, and their spatial distribution (in x, y, and z axes) is encoded to form pose-aware features. Subsequently, we maximize semantic consistency for a pair of pose-text features following a CLIP-based contrastive learning paradigm. Furthermore, a coarse-to-fine mesh regressor is designed, which is capable of effectively querying joint-aware cues from the feature pyramid. Extensive experiments on several public hand benchmarks show that the proposed model attains a significantly faster inference speed while achieving state-of-the-art performance compared to methods utilizing the similar scale backbone.
Distill to Delete: Unlearning in Graph Networks with Knowledge Distillation
Abstract
Graph unlearning has emerged as a pivotal method to delete information from a pre-trained graph neural network (GNN). One may delete nodes, a class of nodes, edges, or a class of edges. An unlearning method enables the GNN model to comply with data protection regulations (i.e., the right to be forgotten), adapt to evolving data distributions, and reduce the GPU-hours carbon footprint by avoiding repetitive retraining. Existing partitioning and aggregation-based methods have limitations due to their poor handling of local graph dependencies and additional overhead costs. More recently, GNNDelete offered a model-agnostic approach that alleviates some of these issues. Our work takes a novel approach to address these challenges in graph unlearning through knowledge distillation, as it distills to delete in GNN (D2DGN). It is a model-agnostic distillation framework where the complete graph knowledge is divided and marked for retention and deletion. It performs distillation with response-based soft targets and feature-based node embedding while minimizing KL divergence. The unlearned model effectively removes the influence of deleted graph elements while preserving knowledge about the retained graph elements. D2DGN surpasses the performance of existing methods when evaluated on various real-world graph datasets by up to $43.1\%$ (AUC) in edge and node unlearning tasks. Other notable advantages include better efficiency, better performance in removing target elements, preservation of performance for the retained elements, and zero overhead costs. Notably, our D2DGN surpasses the state-of-the-art GNNDelete in AUC by $2.4\%$, improves membership inference ratio by $+1.3$, requires $10.2\times10^6$ fewer FLOPs per forward pass and up to $\mathbf{3.2}\times$ faster.
Matrix Multiplication Verification Using Coding Theory
Authors: Huck Bennett, Karthik Gajulapalli, Alexander Golovnev, Philip G. Warton
Abstract
We study the Matrix Multiplication Verification Problem (MMV) where the goal is, given three $n \times n$ matrices $A$, $B$, and $C$ as input, to decide whether $AB = C$. A classic randomized algorithm by Freivalds (MFCS, 1979) solves MMV in $\widetilde{O}(n^2)$ time, and a longstanding challenge is to (partially) derandomize it while still running in faster than matrix multiplication time (i.e., in $o(n^{\omega})$ time). To that end, we give two algorithms for MMV in the case where $AB - C$ is sparse. Specifically, when $AB - C$ has at most $O(n^{\delta})$ non-zero entries for a constant $0 \leq \delta < 2$, we give (1) a deterministic $O(n^{\omega - \varepsilon})$-time algorithm for constant $\varepsilon = \varepsilon(\delta) > 0$, and (2) a randomized $\widetilde{O}(n^2)$-time algorithm using $\delta/2 \cdot \log_2 n + O(1)$ random bits. The former algorithm is faster than the deterministic algorithm of K\"{u}nnemann (ESA, 2018) when $\delta \geq 1.056$, and the latter algorithm uses fewer random bits than the algorithm of Kimbrel and Sinha (IPL, 1993), which runs in the same time and uses $\log_2 n + O(1)$ random bits (in turn fewer than Freivalds's algorithm). We additionally study the complexity of MMV. We first show that all algorithms in a natural class of deterministic linear algebraic algorithms for MMV (including ours) require $\Omega(n^{\omega})$ time. We also show a barrier to proving a super-quadratic running time lower bound for matrix multiplication (and hence MMV) under the Strong Exponential Time Hypothesis (SETH). Finally, we study relationships between natural variants and special cases of MMV (with respect to deterministic $\widetilde{O}(n^2)$-time reductions).
Hierarchical Network Data Analytics Framework for B5G Network Automation: Design and Implementation
Authors: Youbin Jeon, Sangheon Pack
Subjects: Networking and Internet Architecture (cs.NI); Machine Learning (cs.LG); Performance (cs.PF)
Abstract
5G introduced modularized network functions (NFs) to support emerging services in a more flexible and elastic manner. To mitigate the complexity in such modularized NF management, automated network operation and management are indispensable, and thus the 3rd generation partnership project (3GPP) has introduced a network data analytics function (NWDAF). However, a conventional NWDAF needs to conduct both inference and training tasks, and thus it is difficult to provide the analytics results to NFs in a timely manner for an increased number of analytics requests. In this article, we propose a hierarchical network data analytics framework (H-NDAF) where inference tasks are distributed to multiple leaf NWDAFs and training tasks are conducted at the root NWDAF. Extensive simulation results using open-source software (i.e., free5GC) demonstrate that H-NDAF can provide sufficiently accurate analytics and faster analytics provision time compared to the conventional NWDAF.
DeepPCR: Parallelizing Sequential Operations in Neural Networks
Abstract
Parallelization techniques have become ubiquitous for accelerating inference and training of deep neural networks. Despite this, several operations are still performed in a sequential manner. For instance, the forward and backward passes are executed layer-by-layer, and the output of diffusion models is produced by applying a sequence of denoising steps. This sequential approach results in a computational cost proportional to the number of steps involved, presenting a potential bottleneck as the number of steps increases. In this work, we introduce DeepPCR, a novel algorithm which parallelizes typically sequential operations used in inference and training of neural networks. DeepPCR is based on interpreting a sequence of $L$ steps as the solution of a specific system of equations, which we recover using the Parallel Cyclic Reduction algorithm. This reduces the complexity of computing the sequential operations from $\mathcal{O}(L)$ to $\mathcal{O}(\log_2L)$, thus yielding a speedup for large $L$. To verify the theoretical lower complexity of the algorithm, and to identify regimes for speedup, we test the effectiveness of DeepPCR in parallelizing the forward and backward pass in multi-layer perceptrons, and reach speedups of up to $30\times$ for forward and $200\times$ for backward pass. We additionally showcase the flexibility of DeepPCR by parallelizing training of ResNets with as many as 1024 layers, and generation in diffusion models, enabling up to $7\times$ faster training and $11\times$ faster generation, respectively, when compared to the sequential approach.
Efficient Supervised Training of Audio Transformers for Music Representation Learning
Authors: Pablo Alonso-Jiménez, Xavier Serra, Dmitry Bogdanov
Subjects: Sound (cs.SD); Audio and Speech Processing (eess.AS)
Abstract
In this work, we address music representation learning using convolution-free transformers. We build on top of existing spectrogram-based audio transformers such as AST and train our models on a supervised task using patchout training similar to PaSST. In contrast to previous works, we study how specific design decisions affect downstream music tagging tasks instead of focusing on the training task. We assess the impact of initializing the models with different pre-trained weights, using various input audio segment lengths, using learned representations from different blocks and tokens of the transformer for downstream tasks, and applying patchout at inference to speed up feature extraction. We find that 1) initializing the model from ImageNet or AudioSet weights and using longer input segments are beneficial both for the training and downstream tasks, 2) the best representations for the considered downstream tasks are located in the middle blocks of the transformer, and 3) using patchout at inference allows faster processing than our convolutional baselines while maintaining superior performance. The resulting models, MAEST, are publicly available and obtain the best performance among open models in music tagging tasks.
HOI4ABOT: Human-Object Interaction Anticipation for Human Intention Reading Collaborative roBOTs
Authors: Esteve Valls Mascaro, Daniel Sliwowski, Dongheui Lee
Subjects: Computer Vision and Pattern Recognition (cs.CV); Robotics (cs.RO)
Abstract
Robots are becoming increasingly integrated into our lives, assisting us in various tasks. To ensure effective collaboration between humans and robots, it is essential that they understand our intentions and anticipate our actions. In this paper, we propose a Human-Object Interaction (HOI) anticipation framework for collaborative robots. We propose an efficient and robust transformer-based model to detect and anticipate HOIs from videos. This enhanced anticipation empowers robots to proactively assist humans, resulting in more efficient and intuitive collaborations. Our model outperforms state-of-the-art results in HOI detection and anticipation in VidHOI dataset with an increase of 1.76% and 1.04% in mAP respectively while being 15.4 times faster. We showcase the effectiveness of our approach through experimental results in a real robot, demonstrating that the robot's ability to anticipate HOIs is key for better Human-Robot Interaction. More information can be found on our project webpage: https://evm7.github.io/HOI4ABOT_page/
Perceiving Extrinsic Contacts from Touch Improves Learning Insertion Policies
Authors: Carolina Higuera, Joseph Ortiz, Haozhi Qi, Luis Pineda, Byron Boots, Mustafa Mukadam
Abstract
Robotic manipulation tasks such as object insertion typically involve interactions between object and environment, namely extrinsic contacts. Prior work on Neural Contact Fields (NCF) use intrinsic tactile sensing between gripper and object to estimate extrinsic contacts in simulation. However, its effectiveness and utility in real-world tasks remains unknown. In this work, we improve NCF to enable sim-to-real transfer and use it to train policies for mug-in-cupholder and bowl-in-dishrack insertion tasks. We find our model NCF-v2, is capable of estimating extrinsic contacts in the real-world. Furthermore, our insertion policy with NCF-v2 outperforms policies without it, achieving 33% higher success and 1.36x faster execution on mug-in-cupholder, and 13% higher success and 1.27x faster execution on bowl-in-dishrack.
DreamGaussian: Generative Gaussian Splatting for Efficient 3D Content Creation
Authors: Jiaxiang Tang, Jiawei Ren, Hang Zhou, Ziwei Liu, Gang Zeng
Subjects: Computer Vision and Pattern Recognition (cs.CV)
Abstract
Recent advances in 3D content creation mostly leverage optimization-based 3D generation via score distillation sampling (SDS). Though promising results have been exhibited, these methods often suffer from slow per-sample optimization, limiting their practical usage. In this paper, we propose DreamGaussian, a novel 3D content generation framework that achieves both efficiency and quality simultaneously. Our key insight is to design a generative 3D Gaussian Splatting model with companioned mesh extraction and texture refinement in UV space. In contrast to the occupancy pruning used in Neural Radiance Fields, we demonstrate that the progressive densification of 3D Gaussians converges significantly faster for 3D generative tasks. To further enhance the texture quality and facilitate downstream applications, we introduce an efficient algorithm to convert 3D Gaussians into textured meshes and apply a fine-tuning stage to refine the details. Extensive experiments demonstrate the superior efficiency and competitive generation quality of our proposed approach. Notably, DreamGaussian produces high-quality textured meshes in just 2 minutes from a single-view image, achieving approximately 10 times acceleration compared to existing methods.
Visual In-Context Learning for Few-Shot Eczema Segmentation
Abstract
Automated diagnosis of eczema from digital camera images is crucial for developing applications that allow patients to self-monitor their recovery. An important component of this is the segmentation of eczema region from such images. Current methods for eczema segmentation rely on deep neural networks such as convolutional (CNN)-based U-Net or transformer-based Swin U-Net. While effective, these methods require high volume of annotated data, which can be difficult to obtain. Here, we investigate the capabilities of visual in-context learning that can perform few-shot eczema segmentation with just a handful of examples and without any need for retraining models. Specifically, we propose a strategy for applying in-context learning for eczema segmentation with a generalist vision model called SegGPT. When benchmarked on a dataset of annotated eczema images, we show that SegGPT with just 2 representative example images from the training dataset performs better (mIoU: 36.69) than a CNN U-Net trained on 428 images (mIoU: 32.60). We also discover that using more number of examples for SegGPT may in fact be harmful to its performance. Our result highlights the importance of visual in-context learning in developing faster and better solutions to skin imaging tasks. Our result also paves the way for developing inclusive solutions that can cater to minorities in the demographics who are typically heavily under-represented in the training data.
HyperPPO: A scalable method for finding small policies for robotic control
Authors: Shashank Hegde, Zhehui Huang, Gaurav S. Sukhatme
Abstract
Models with fewer parameters are necessary for the neural control of memory-limited, performant robots. Finding these smaller neural network architectures can be time-consuming. We propose HyperPPO, an on-policy reinforcement learning algorithm that utilizes graph hypernetworks to estimate the weights of multiple neural architectures simultaneously. Our method estimates weights for networks that are much smaller than those in common-use networks yet encode highly performant policies. We obtain multiple trained policies at the same time while maintaining sample efficiency and provide the user the choice of picking a network architecture that satisfies their computational constraints. We show that our method scales well - more training resources produce faster convergence to higher-performing architectures. We demonstrate that the neural policies estimated by HyperPPO are capable of decentralized control of a Crazyflie2.1 quadrotor. Website: https://sites.google.com/usc.edu/hyperppo
Keyword: mobile
Breamy: An augmented reality mHealth prototype for surgical decision-making in breast cancer
Authors: Niki Najafi, Miranda Addie, Sarkis Meterissian, Marta Kersten-Oertel
Abstract
In 2020, according to WHO, breast cancer affected 2.3 million women worldwide, resulting in 685,000 fatalities. By the end of the year, approximately 7.8 million women worldwide had survived their breast cancer making it the most widespread form of cancer globally. Surgical treatment decisions, including choosing between oncoplastic options, often require quick decision-making within an 8-week time frame. However, many women lack the necessary knowledge and preparation for making such complex informed decisions. Anxiety and unsatisfactory outcomes can result from inadequate decision-making processes, leading to complications and the need for revision surgeries. Shared decision-making and personalized decision aids have shown positive effects on patient satisfaction and treatment outcomes. This paper introduces Breamy, a prototype mobile health (mHealth) application that utilizes augmented reality (AR) technology to assist breast cancer patients in making informed decisions. The app provides 3D visualizations of different oncoplastic procedures, aiming to improve confidence in surgical decision-making, reduce decisional regret, and enhance patient well-being after surgery. To determine the perception of the usefulness of Breamy, we collected data from 166 participants through an online survey. The results suggest that Breamy has the potential to reduce patient's anxiety levels and assist them during the decision-making process.
Dynamics of Ideological Biases of Social Media Users
Authors: Mohammed Shahid Modi, James Flamino, Boleslaw K. Szymanski
Subjects: Social and Information Networks (cs.SI); Computers and Society (cs.CY)
Abstract
Humanity for centuries has perfected skills of interpersonal interactions and evolved patterns that enable people to detect lies and deceiving behavior of others in face-to-face settings. Unprecedented growth of people's access to mobile phones and social media raises an important question: How does this new technology influence people's interactions and support the use of traditional patterns? In this paper, we answer this question for homophily driven patterns in social media. In our previous studies, we found that, on a university campus, changes in student opinions were driven by the desire to hold popular opinions. Here, we demonstrate that the evolution of online platform-wide opinion groups is driven by the same desire. We focus on two social media: Twitter and Parler, on which we tracked the political biases of their users. On Parler, an initially stable group of right-biased users evolved into a permanent right-leaning echo chamber dominating weaker, transient groups of members with opposing political biases. In contrast, on Twitter, the initial presence of two large opposing bias groups led to the evolution of a bimodal bias distribution, with a high degree of polarization. We capture the movement of users from the initial to final bias groups during the tracking period. We also show that user choices are influenced by side-effects of homophily. The users entering the platform attempt to find a sufficiently large group whose members hold political bias within the range sufficiently close to the new user's bias. If successful, they stabilize their bias and become a permanent member of the group. Otherwise, they leave the platform. We believe that the dynamics of users uncovered in this paper create a foundation for technical solutions supporting social groups on social media and socially aware networks.
DynaCon: Dynamic Robot Planner with Contextual Awareness via LLMs
Authors: Gyeongmin Kim, Taehyeon Kim, Shyam Sundar Kannan, Vishnunandan L. N. Venkatesh, Donghan Kim, Byung-Cheol Min
Abstract
Mobile robots often rely on pre-existing maps for effective path planning and navigation. However, when these maps are unavailable, particularly in unfamiliar environments, a different approach become essential. This paper introduces DynaCon, a novel system designed to provide mobile robots with contextual awareness and dynamic adaptability during navigation, eliminating the reliance of traditional maps. DynaCon integrates real-time feedback with an object server, prompt engineering, and navigation modules. By harnessing the capabilities of Large Language Models (LLMs), DynaCon not only understands patterns within given numeric series but also excels at categorizing objects into matched spaces. This facilitates dynamic path planner imbued with contextual awareness. We validated the effectiveness of DynaCon through an experiment where a robot successfully navigated to its goal using reasoning. Source code and experiment videos for this work can be found at: https://sites.google.com/view/dynacon.
Nondestructive chicken egg fertility detection using CNN-transfer learning algorithms
Authors: Shoffan Saifullah, Rafal Drezewski, Anton Yudhana, Andri Pranolo, Wilis Kaswijanti, Andiko Putro Suryotomo, Seno Aji Putra, Alin Khaliduzzaman, Anton Satria Prabuwono, Nathalie Japkowicz
Subjects: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI); Image and Video Processing (eess.IV)
Abstract
This study explored the application of CNN-Transfer Learning for nondestructive chicken egg fertility detection for precision poultry hatchery practices. Four models, VGG16, ResNet50, InceptionNet, and MobileNet, were trained and evaluated on a dataset (200 single egg images) using augmented images (rotation, flip, scale, translation, and reflection). Although the training results demonstrated that all models achieved high accuracy, indicating their ability to accurately learn and classify chicken eggs' fertility state, when evaluated on the testing set, variations in accuracy and performance were observed. InceptionNet exhibited the best overall performance, accurately classifying fertile and non-fertile eggs. It demonstrated excellent performance in both training and testing sets in all parameters of the evaluation metrics. In testing set, it achieved an accuracy of 0.98, a sensitivity of 1 for detecting fertile eggs, and a specificity of 0.96 for identifying non-fertile eggs. The higher performance is attributed to its unique architecture efficiently capturing features at different scales leading to improved accuracy and robustness. Further optimization and fine-tuning of the models might necessary to address the limitations in accurately detecting fertile and non-fertile eggs in case of other models. This study highlighted the potential of CNN-Transfer Learning for nondestructive fertility detection and emphasizes the need for further research to enhance the models' capabilities and ensure accurate classification.
Semantic Scene Difference Detection in Daily Life Patroling by Mobile Robots using Pre-Trained Large-Scale Vision-Language Model
Abstract
It is important for daily life support robots to detect changes in their environment and perform tasks. In the field of anomaly detection in computer vision, probabilistic and deep learning methods have been used to calculate the image distance. These methods calculate distances by focusing on image pixels. In contrast, this study aims to detect semantic changes in the daily life environment using the current development of large-scale vision-language models. Using its Visual Question Answering (VQA) model, we propose a method to detect semantic changes by applying multiple questions to a reference image and a current image and obtaining answers in the form of sentences. Unlike deep learning-based methods in anomaly detection, this method does not require any training or fine-tuning, is not affected by noise, and is sensitive to semantic state changes in the real world. In our experiments, we demonstrated the effectiveness of this method by applying it to a patrol task in a real-life environment using a mobile robot, Fetch Mobile Manipulator. In the future, it may be possible to add explanatory power to changes in the daily life environment through spoken language.
Review of Machine Learning Methods for Additive Manufacturing of Functionally Graded Materials
Authors: Mohammad Karimzadeh, Aleksandar Vakanski, Fei Xu, Xinchang Zhang
Abstract
Additive manufacturing has revolutionized the manufacturing of complex parts by enabling direct material joining and offers several advantages such as cost-effective manufacturing of complex parts, reducing manufacturing waste, and opening new possibilities for manufacturing automation. One group of materials for which additive manufacturing holds great potential for enhancing component performance and properties is Functionally Graded Materials (FGMs). FGMs are advanced composite materials that exhibit smoothly varying properties making them desirable for applications in aerospace, automobile, biomedical, and defense industries. Such composition differs from traditional composite materials, since the location-dependent composition changes gradually in FGMs, leading to enhanced properties. Recently, machine learning techniques have emerged as a promising means for fabrication of FGMs through optimizing processing parameters, improving product quality, and detecting manufacturing defects. This paper first provides a brief literature review of works related to FGM fabrication, followed by reviewing works on employing machine learning in additive manufacturing, Afterward, we provide an overview of published works in the literature related to the application of machine learning methods in Directed Energy Deposition and for fabrication of FGMs.
Keyword: pruning
Discovering Utility-driven Interval Rules
Authors: Chunkai Zhang, Maohua Lyu, Huaijin Hao, Wensheng Gan, Philip S. Yu
Abstract
For artificial intelligence, high-utility sequential rule mining (HUSRM) is a knowledge discovery method that can reveal the associations between events in the sequences. Recently, abundant methods have been proposed to discover high-utility sequence rules. However, the existing methods are all related to point-based sequences. Interval events that persist for some time are common. Traditional interval-event sequence knowledge discovery tasks mainly focus on pattern discovery, but patterns cannot reveal the correlation between interval events well. Moreover, the existing HUSRM algorithms cannot be directly applied to interval-event sequences since the relation in interval-event sequences is much more intricate than those in point-based sequences. In this work, we propose a utility-driven interval rule mining (UIRMiner) algorithm that can extract all utility-driven interval rules (UIRs) from the interval-event sequence database to solve the problem. In UIRMiner, we first introduce a numeric encoding relation representation, which can save much time on relation computation and storage on relation representation. Furthermore, to shrink the search space, we also propose a complement pruning strategy, which incorporates the utility upper bound with the relation. Finally, plentiful experiments implemented on both real-world and synthetic datasets verify that UIRMiner is an effective and efficient algorithm.
Rethinking Domain Generalization: Discriminability and Generalizability
Abstract
Domain generalization (DG) endeavors to develop robust models that possess strong generalizability while preserving excellent discriminability. Nonetheless, pivotal DG techniques tend to improve the feature generalizability by learning domain-invariant representations, inadvertently overlooking the feature discriminability. On the one hand, the simultaneous attainment of generalizability and discriminability of features presents a complex challenge, often entailing inherent contradictions. This challenge becomes particularly pronounced when domain-invariant features manifest reduced discriminability owing to the inclusion of unstable factors, \emph{i.e.,} spurious correlations. On the other hand, prevailing domain-invariant methods can be categorized as category-level alignment, susceptible to discarding indispensable features possessing substantial generalizability and narrowing intra-class variations. To surmount these obstacles, we rethink DG from a new perspective that concurrently imbues features with formidable discriminability and robust generalizability, and present a novel framework, namely, Discriminative Microscopic Distribution Alignment (DMDA). DMDA incorporates two core components: Selective Channel Pruning~(SCP) and Micro-level Distribution Alignment (MDA). Concretely, SCP attempts to curtail redundancy within neural networks, prioritizing stable attributes conducive to accurate classification. This approach alleviates the adverse effect of spurious domain invariance and amplifies the feature discriminability. Besides, MDA accentuates micro-level alignment within each class, going beyond mere category-level alignment. This strategy accommodates sufficient generalizable features and facilitates within-class variations. Extensive experiments on four benchmark datasets corroborate the efficacy of our method.
DreamGaussian: Generative Gaussian Splatting for Efficient 3D Content Creation
Authors: Jiaxiang Tang, Jiawei Ren, Hang Zhou, Ziwei Liu, Gang Zeng
Subjects: Computer Vision and Pattern Recognition (cs.CV)
Abstract
Recent advances in 3D content creation mostly leverage optimization-based 3D generation via score distillation sampling (SDS). Though promising results have been exhibited, these methods often suffer from slow per-sample optimization, limiting their practical usage. In this paper, we propose DreamGaussian, a novel 3D content generation framework that achieves both efficiency and quality simultaneously. Our key insight is to design a generative 3D Gaussian Splatting model with companioned mesh extraction and texture refinement in UV space. In contrast to the occupancy pruning used in Neural Radiance Fields, we demonstrate that the progressive densification of 3D Gaussians converges significantly faster for 3D generative tasks. To further enhance the texture quality and facilitate downstream applications, we introduce an efficient algorithm to convert 3D Gaussians into textured meshes and apply a fine-tuning stage to refine the details. Extensive experiments demonstrate the superior efficiency and competitive generation quality of our proposed approach. Notably, DreamGaussian produces high-quality textured meshes in just 2 minutes from a single-view image, achieving approximately 10 times acceleration compared to existing methods.
Keyword: diffusion
A priori and a posteriori error analysis for a VEM discretization of the convection-diffusion eigenvalue problem
Authors: Danilo Amigo, Felipe Lepe, Gonzalo Rivera
Abstract
In this paper we propose and analyze a virtual element method for the two dimensional non-symmetric diffusion-convection eigenvalue problem in order to derive a priori and a posteriori error estimates. Under the classic assumptions of the meshes, and with the aid of the classic theory of compact operators, we prove error estimates for the eigenvalues and eigenfunctions. Also, we develop an a posteriori error estimator which, in one hand, results to be reliable and on the other, with standard bubble functions arguments, also results to be efficient. We test our method on domains where the complex eigenfunctions are not sufficiently regular, in order to assess the performance of the estimator that we compare with the uniform refinement given by the a priori analysis
Compositional Sculpting of Iterative Generative Processes
Authors: Timur Garipov, Sebastiaan De Peuter, Ge Yang, Vikas Garg, Samuel Kaski, Tommi Jaakkola
Abstract
High training costs of generative models and the need to fine-tune them for specific tasks have created a strong interest in model reuse and composition. A key challenge in composing iterative generative processes, such as GFlowNets and diffusion models, is that to realize the desired target distribution, all steps of the generative process need to be coordinated, and satisfy delicate balance conditions. In this work, we propose Compositional Sculpting: a general approach for defining compositions of iterative generative processes. We then introduce a method for sampling from these compositions built on classifier guidance. We showcase ways to accomplish compositional sculpting in both GFlowNets and diffusion models. We highlight two binary operations $\unicode{x2014}$ the harmonic mean ($p_1 \otimes p_2$) and the contrast ($p_1 \unicode{x25D1}\,p_2$) between pairs, and the generalization of these operations to multiple component distributions. We offer empirical results on image and molecular generation tasks.
DiffGAN-F2S: Symmetric and Efficient Denoising Diffusion GANs for Structural Connectivity Prediction from Brain fMRI
Authors: Qiankun Zuo, Ruiheng Li, Yi Di, Hao Tian, Changhong Jing, Xuhang Chen, Shuqiang Wang
Subjects: Computer Vision and Pattern Recognition (cs.CV); Image and Video Processing (eess.IV)
Abstract
Mapping from functional connectivity (FC) to structural connectivity (SC) can facilitate multimodal brain network fusion and discover potential biomarkers for clinical implications. However, it is challenging to directly bridge the reliable non-linear mapping relations between SC and functional magnetic resonance imaging (fMRI). In this paper, a novel diffusision generative adversarial network-based fMRI-to-SC (DiffGAN-F2S) model is proposed to predict SC from brain fMRI in an end-to-end manner. To be specific, the proposed DiffGAN-F2S leverages denoising diffusion probabilistic models (DDPMs) and adversarial learning to efficiently generate high-fidelity SC through a few steps from fMRI. By designing the dual-channel multi-head spatial attention (DMSA) and graph convolutional modules, the symmetric graph generator first captures global relations among direct and indirect connected brain regions, then models the local brain region interactions. It can uncover the complex mapping relations between fMRI and structural connectivity. Furthermore, the spatially connected consistency loss is devised to constrain the generator to preserve global-local topological information for accurate intrinsic SC prediction. Testing on the public Alzheimer's Disease Neuroimaging Initiative (ADNI) dataset, the proposed model can effectively generate empirical SC-preserved connectivity from four-dimensional imaging data and shows superior performance in SC prediction compared with other related models. Furthermore, the proposed model can identify the vast majority of important brain regions and connections derived from the empirical method, providing an alternative way to fuse multimodal brain networks and analyze clinical disease.
Object Motion Guided Human Motion Synthesis
Authors: Jiaman Li, Jiajun Wu, C. Karen Liu
Subjects: Computer Vision and Pattern Recognition (cs.CV)
Abstract
Modeling human behaviors in contextual environments has a wide range of applications in character animation, embodied AI, VR/AR, and robotics. In real-world scenarios, humans frequently interact with the environment and manipulate various objects to complete daily tasks. In this work, we study the problem of full-body human motion synthesis for the manipulation of large-sized objects. We propose Object MOtion guided human MOtion synthesis (OMOMO), a conditional diffusion framework that can generate full-body manipulation behaviors from only the object motion. Since naively applying diffusion models fails to precisely enforce contact constraints between the hands and the object, OMOMO learns two separate denoising processes to first predict hand positions from object motion and subsequently synthesize full-body poses based on the predicted hand positions. By employing the hand positions as an intermediate representation between the two denoising processes, we can explicitly enforce contact constraints, resulting in more physically plausible manipulation motions. With the learned model, we develop a novel system that captures full-body human manipulation motions by simply attaching a smartphone to the object being manipulated. Through extensive experiments, we demonstrate the effectiveness of our proposed pipeline and its ability to generalize to unseen objects. Additionally, as high-quality human-object interaction datasets are scarce, we collect a large-scale dataset consisting of 3D object geometry, object motion, and human motion. Our dataset contains human-object interaction motion for 15 objects, with a total duration of approximately 10 hours.
DeepPCR: Parallelizing Sequential Operations in Neural Networks
Abstract
Parallelization techniques have become ubiquitous for accelerating inference and training of deep neural networks. Despite this, several operations are still performed in a sequential manner. For instance, the forward and backward passes are executed layer-by-layer, and the output of diffusion models is produced by applying a sequence of denoising steps. This sequential approach results in a computational cost proportional to the number of steps involved, presenting a potential bottleneck as the number of steps increases. In this work, we introduce DeepPCR, a novel algorithm which parallelizes typically sequential operations used in inference and training of neural networks. DeepPCR is based on interpreting a sequence of $L$ steps as the solution of a specific system of equations, which we recover using the Parallel Cyclic Reduction algorithm. This reduces the complexity of computing the sequential operations from $\mathcal{O}(L)$ to $\mathcal{O}(\log_2L)$, thus yielding a speedup for large $L$. To verify the theoretical lower complexity of the algorithm, and to identify regimes for speedup, we test the effectiveness of DeepPCR in parallelizing the forward and backward pass in multi-layer perceptrons, and reach speedups of up to $30\times$ for forward and $200\times$ for backward pass. We additionally showcase the flexibility of DeepPCR by parallelizing training of ResNets with as many as 1024 layers, and generation in diffusion models, enabling up to $7\times$ faster training and $11\times$ faster generation, respectively, when compared to the sequential approach.
Meshless interface tracking for the simulation of dendrite envelope growth
Authors: Mitja Jančič, Miha Založnik, Gregor Kosec
Abstract
The growth of dendritic grains during solidification is often modelled using the Grain Envelope Model (GEM), in which the envelope of the dendrite is an interface tracked by the Phase Field Interface Capturing (PFIC) method. In the PFIC method, an phase-field equation is solved on a fixed mesh to track the position of the envelope. While being versatile and robust, PFIC introduces certain numerical artefacts. In this work, we present an alternative approach for the solution of the GEM that employs a Meshless (sharp) Interface Tracking (MIT) formulation, which uses direct, artefact-free interface tracking. In the MIT, the envelope (interface) is defined as a moving domain boundary and the interface-tracking nodes are boundary nodes for the diffusion problem solved in the domain. To increase the accuracy of the method for the diffusion-controlled moving-boundary problem, an \h-adaptive spatial discretization is used, thus, the node spacing is refined in the vicinity of the envelope. MIT combines a parametric surface reconstruction, a mesh-free discretization of the parametric surfaces and the space enclosed by them, and a high-order approximation of the partial differential operators and of the solute concentration field using radial basis functions augmented with monomials. The proposed method is demonstrated on a two-dimensional \h-adaptive solution of the diffusive growth of dendrite and evaluated by comparing the results to the PFIC approach. It is shown that MIT can reproduce the results calculated with PFIC, that it is convergent and that it can capture more details in the envelope shape than PFIC with a similar spatial discretization.
Distilling ODE Solvers of Diffusion Models into Smaller Steps
Authors: Sanghwan Kim, Hao Tang, Fisher Yu
Subjects: Computer Vision and Pattern Recognition (cs.CV)
Abstract
Distillation techniques have substantially improved the sampling speed of diffusion models, allowing of the generation within only one step or a few steps. However, these distillation methods require extensive training for each dataset, sampler, and network, which limits their practical applicability. To address this limitation, we propose a straightforward distillation approach, Distilled-ODE solvers (D-ODE solvers), that optimizes the ODE solver rather than training the denoising network. D-ODE solvers are formulated by simply applying a single parameter adjustment to existing ODE solvers. Subsequently, D-ODE solvers with smaller steps are optimized by ODE solvers with larger steps through distillation over a batch of samples. Our comprehensive experiments indicate that D-ODE solvers outperform existing ODE solvers, including DDIM, PNDM, DPM-Solver, DEIS, and EDM, especially when generating samples with fewer steps. Our method incur negligible computational overhead compared to previous distillation techniques, enabling simple and rapid integration with previous samplers. Qualitative analysis further shows that D-ODE solvers enhance image quality while preserving the sampling trajectory of ODE solvers.
KV Inversion: KV Embeddings Learning for Text-Conditioned Real Image Action Editing
Authors: Jiancheng Huang, Yifan Liu, Jin Qin, Shifeng Chen
Subjects: Computer Vision and Pattern Recognition (cs.CV)
Abstract
Text-conditioned image editing is a recently emerged and highly practical task, and its potential is immeasurable. However, most of the concurrent methods are unable to perform action editing, i.e. they can not produce results that conform to the action semantics of the editing prompt and preserve the content of the original image. To solve the problem of action editing, we propose KV Inversion, a method that can achieve satisfactory reconstruction performance and action editing, which can solve two major problems: 1) the edited result can match the corresponding action, and 2) the edited object can retain the texture and identity of the original real image. In addition, our method does not require training the Stable Diffusion model itself, nor does it require scanning a large-scale dataset to perform time-consuming training.
Keyword: adaptive
Enhancing Cross-Category Learning in Recommendation Systems with Multi-Layer Embedding Training
Authors: Zihao Deng, Benjamin Ghaemmaghami, Ashish Kumar Singh, Benjamin Cho, Leo Orshansky, Mattan Erez, Michael Orshansky
Abstract
Modern DNN-based recommendation systems rely on training-derived embeddings of sparse features. Input sparsity makes obtaining high-quality embeddings for rarely-occurring categories harder as their representations are updated infrequently. We demonstrate a training-time technique to produce superior embeddings via effective cross-category learning and theoretically explain its surprising effectiveness. The scheme, termed the multi-layer embeddings training (MLET), trains embeddings using factorization of the embedding layer, with an inner dimension higher than the target embedding dimension. For inference efficiency, MLET converts the trained two-layer embedding into a single-layer one thus keeping inference-time model size unchanged. Empirical superiority of MLET is puzzling as its search space is not larger than that of the single-layer embedding. The strong dependence of MLET on the inner dimension is even more surprising. We develop a theory that explains both of these behaviors by showing that MLET creates an adaptive update mechanism modulated by the singular vectors of embeddings. When tested on multiple state-of-the-art recommendation models for click-through rate (CTR) prediction tasks, MLET consistently produces better models, especially for rare items. At constant model quality, MLET allows embedding dimension, and model size, reduction by up to 16x, and 5.8x on average, across the models.
Improving Adaptive Online Learning Using Refined Discretization
Authors: Zhiyu Zhang, Heng Yang, Ashok Cutkosky, Ioannis Ch. Paschalidis
Abstract
We study unconstrained Online Linear Optimization with Lipschitz losses. The goal is to simultaneously achieve ($i$) second order gradient adaptivity; and ($ii$) comparator norm adaptivity also known as "parameter freeness" in the literature. Existing regret bounds (Cutkosky and Orabona, 2018; Mhammedi and Koolen, 2020; Jacobsen and Cutkosky, 2022) have the suboptimal $O(\sqrt{V_T\log V_T})$ dependence on the gradient variance $V_T$, while the present work improves it to the optimal rate $O(\sqrt{V_T})$ using a novel continuous-time-inspired algorithm, without any impractical doubling trick. This result can be extended to the setting with unknown Lipschitz constant, eliminating the range ratio problem from prior works (Mhammedi and Koolen, 2020). Concretely, we first show that the aimed simultaneous adaptivity can be achieved fairly easily in a continuous time analogue of the problem, where the environment is modeled by an arbitrary continuous semimartingale. Then, our key innovation is a new discretization argument that preserves such adaptivity in the discrete time adversarial setting. This refines a non-gradient-adaptive discretization argument from (Harvey et al., 2023), both algorithmically and analytically, which could be of independent interest.
A review of variable-pitch propellers and their control strategies in aerospace systems
Abstract
The relentless pursuit of aircraft flight efficiency has thrust variable-pitch propeller technology into the forefront of aviation innovation. This technology, rooted in the ancient power unit of propellers, has found renewed significance, particularly in the realms of unmanned aerial vehicles and urban air mobility. This underscores the profound interplay between visionary aviation concepts and the enduring utility of propellers. Variable-pitch propellers are poised to be pivotal in shaping the future of human aviation, offering benefits such as extended endurance, enhanced maneuverability, improved fuel economy, and prolonged engine life. However, with additional capabilities come new technical challenges. The development of an online adaptive control of variable-pitch propellers that does not depend on an accurate dynamic model stands as a critical imperative. Therefore, a comprehensive review and forward-looking analysis of this technology is warranted. This paper introduces the development background of variable-pitch aviation propeller technology, encompassing diverse pitch angle adjustment schemes and their integration with various engine types. It places a central focus on the latest research frontiers and emerging directions in pitch control strategies. Lastly, it delves into the research domain of constant speed pitch control, articulating the three main challenges confronting this technology: inadequacies in system modeling, the intricacies of propeller-engine compatibility, and the impact of external, time-varying factors. By shedding light on these multifaceted aspects of variable-pitch propeller technology, this paper serves as a resource for aviation professionals and researchers navigating the intricate landscape of future aircraft development.
Differentiable Robot Neural Distance Function for Adaptive Grasp Synthesis on a Unified Robotic Arm-Hand System
Abstract
Grasping is a fundamental skill for robots to interact with their environment. While grasp execution requires coordinated movement of the hand and arm to achieve a collision-free and secure grip, many grasp synthesis studies address arm and hand motion planning independently, leading to potentially unreachable grasps in practical settings. The challenge of determining integrated arm-hand configurations arises from its computational complexity and high-dimensional nature. We address this challenge by presenting a novel differentiable robot neural distance function. Our approach excels in capturing intricate geometry across various joint configurations while preserving differentiability. This innovative representation proves instrumental in efficiently addressing downstream tasks with stringent contact constraints. Leveraging this, we introduce an adaptive grasp synthesis framework that exploits the full potential of the unified arm-hand system for diverse grasping tasks. Our neural joint space distance function achieves an 84.7% error reduction compared to baseline methods. We validated our approaches on a unified robotic arm-hand system that consists of a 7-DoF robot arm and a 16-DoF multi-fingered robotic hand. Results demonstrate that our approach empowers this high-DoF system to generate and execute various arm-hand grasp configurations that adapt to the size of the target objects while ensuring whole-body movements to be collision-free.
ModuLoRA: Finetuning 3-Bit LLMs on Consumer GPUs by Integrating with Modular Quantizers
Authors: Junjie Yin, Jiahao Dong, Yingheng Wang, Christopher De Sa, Volodymyr Kuleshov
Abstract
We propose a memory-efficient finetuning algorithm for large language models (LLMs) that supports finetuning LLMs with 65B parameters in 3-bit or 4-bit precision on as little as one 48GB GPU. Our method, modular low-rank adaptation (ModuLoRA), integrates any user-specified weight quantizer with finetuning via low-rank adapters (LoRAs). Our approach relies on a simple quantization-agnostic backward pass that adaptively materializes low-precision LLM weights from a custom black-box quantization module. This approach enables finetuning 3-bit LLMs for the first time--leveraging state-of-the-art 3-bit OPTQ quantization often outperforms finetuning that relies on less sophisticated 4-bit and 8-bit methods. In our experiments, ModuLoRA attains competitive performance on text classification, natural language infernece, and instruction following tasks using significantly less memory than existing approaches, and we also surpass the state-of-the-art ROUGE score on a popular summarization task. We release ModuLoRA together with a series of low-precision models--including the first family of 3-bit instruction following Alpaca LLMs--as part of LLMTOOLS, a user-friendly library for quantizing, running, and finetuning LLMs on consumer GPUs.
Joint Correcting and Refinement for Balanced Low-Light Image Enhancement
Authors: Nana Yu, Hong Shi, Jie Wang, Yahong Han
Subjects: Computer Vision and Pattern Recognition (cs.CV); Image and Video Processing (eess.IV)
Abstract
Low-light image enhancement tasks demand an appropriate balance among brightness, color, and illumination. While existing methods often focus on one aspect of the image without considering how to pay attention to this balance, which will cause problems of color distortion and overexposure etc. This seriously affects both human visual perception and the performance of high-level visual models. In this work, a novel synergistic structure is proposed which can balance brightness, color, and illumination more effectively. Specifically, the proposed method, so-called Joint Correcting and Refinement Network (JCRNet), which mainly consists of three stages to balance brightness, color, and illumination of enhancement. Stage 1: we utilize a basic encoder-decoder and local supervision mechanism to extract local information and more comprehensive details for enhancement. Stage 2: cross-stage feature transmission and spatial feature transformation further facilitate color correction and feature refinement. Stage 3: we employ a dynamic illumination adjustment approach to embed residuals between predicted and ground truth images into the model, adaptively adjusting illumination balance. Extensive experiments demonstrate that the proposed method exhibits comprehensive performance advantages over 21 state-of-the-art methods on 9 benchmark datasets. Furthermore, a more persuasive experiment has been conducted to validate our approach the effectiveness in downstream visual tasks (e.g., saliency detection). Compared to several enhancement models, the proposed method effectively improves the segmentation results and quantitative metrics of saliency detection. The source code will be available at https://github.com/woshiyll/JCRNet.
Context-I2W: Mapping Images to Context-dependent Words for Accurate Zero-Shot Composed Image Retrieval
Abstract
Different from Composed Image Retrieval task that requires expensive labels for training task-specific models, Zero-Shot Composed Image Retrieval (ZS-CIR) involves diverse tasks with a broad range of visual content manipulation intent that could be related to domain, scene, object, and attribute. The key challenge for ZS-CIR tasks is to learn a more accurate image representation that has adaptive attention to the reference image for various manipulation descriptions. In this paper, we propose a novel context-dependent mapping network, named Context-I2W, for adaptively converting description-relevant Image information into a pseudo-word token composed of the description for accurate ZS-CIR. Specifically, an Intent View Selector first dynamically learns a rotation rule to map the identical image to a task-specific manipulation view. Then a Visual Target Extractor further captures local information covering the main targets in ZS-CIR tasks under the guidance of multiple learnable queries. The two complementary modules work together to map an image to a context-dependent pseudo-word token without extra supervision. Our model shows strong generalization ability on four ZS-CIR tasks, including domain conversion, object composition, object manipulation, and attribute manipulation. It obtains consistent and significant performance boosts ranging from 1.88% to 3.60% over the best methods and achieves new state-of-the-art results on ZS-CIR. Our code is available at https://github.com/Pter61/context_i2w.
Adaptive Real-Time Numerical Differentiation with Variable-Rate Forgetting and Exponential Resetting
Authors: Shashank Verma, Brian Lai, Dennis S. Bernstein
Subjects: Systems and Control (eess.SY); Signal Processing (eess.SP)
Abstract
Digital PID control requires a differencing operation to implement the D gain. In order to suppress the effects of noisy data, the traditional approach is to filter the data, where the frequency response of the filter is adjusted manually based on the characteristics of the sensor noise. The present paper considers the case where the characteristics of the sensor noise change over time in an unknown way. This problem is addressed by applying adaptive real-time numerical differentiation based on adaptive input and state estimation (AISE). The contribution of this paper is to extend AISE to include variable-rate forgetting with exponential resetting, which allows AISE to more rapidly respond to changing noise characteristics while enforcing the boundedness of the covariance matrix used in recursive least squares.
Laboratory Automation: Precision Insertion with Adaptive Fingers utilizing Contact through Sliding with Tactile-based Pose Estimation
Abstract
Micro well-plates are commonly used apparatus in chemical and biological experiments that are a few centimeters in thickness with wells in them. The task we aim to solve is to place (insert) them onto a well-plate holder with grooves a few millimeters in height. Our insertion task has the following facets: 1) There is uncertainty in the detection of the position and pose of the well-plate and well-plate holder, 2) the accuracy required is in the order of millimeter to sub-millimeter, 3) the well-plate holder is not fastened, and moves with external force, 4) the groove is shallow, and 5) the width of the groove is small. Addressing these challenges, we developed a) an adaptive finger gripper with accurate detection of finger position (for (1)), b) grasped object pose estimation using tactile sensors (for (1)), c) a method to insert the well-plate into the target holder by sliding the well-plate while maintaining contact with the edge of the holder (for (2-4)), and d) estimating the orientation of the edge and aligning the well-plate so that the holder does not move when maintaining contact with the edge (for (5)). We show a significantly high success rate on the insertion task of the well-plate, even though under added noise. An accompanying video is available at the following link: https://drive.google.com/file/d/1UxyJ3XIxqXPnHcpfw-PYs5T5oYQxoc6i/view?usp=sharing
Cloth2Body: Generating 3D Human Body Mesh from 2D Clothing
Abstract
In this paper, we define and study a new Cloth2Body problem which has a goal of generating 3D human body meshes from a 2D clothing image. Unlike the existing human mesh recovery problem, Cloth2Body needs to address new and emerging challenges raised by the partial observation of the input and the high diversity of the output. Indeed, there are three specific challenges. First, how to locate and pose human bodies into the clothes. Second, how to effectively estimate body shapes out of various clothing types. Finally, how to generate diverse and plausible results from a 2D clothing image. To this end, we propose an end-to-end framework that can accurately estimate 3D body mesh parameterized by pose and shape from a 2D clothing image. Along this line, we first utilize Kinematics-aware Pose Estimation to estimate body pose parameters. 3D skeleton is employed as a proxy followed by an inverse kinematics module to boost the estimation accuracy. We additionally design an adaptive depth trick to align the re-projected 3D mesh better with 2D clothing image by disentangling the effects of object size and camera extrinsic. Next, we propose Physics-informed Shape Estimation to estimate body shape parameters. 3D shape parameters are predicted based on partial body measurements estimated from RGB image, which not only improves pixel-wise human-cloth alignment, but also enables flexible user editing. Finally, we design Evolution-based pose generation method, a skeleton transplanting method inspired by genetic algorithms to generate diverse reasonable poses during inference. As shown by experimental results on both synthetic and real-world data, the proposed framework achieves state-of-the-art performance and can effectively recover natural and diverse 3D body meshes from 2D images that align well with clothing.
GAMMA: Generalizable Articulation Modeling and Manipulation for Articulated Objects
Authors: Qiaojun Yu, Junbo Wang, Wenhai Liu, Ce Hao, Liu Liu, Lin Shao, Weiming Wang, Cewu Lu
Subjects: Robotics (cs.RO); Computer Vision and Pattern Recognition (cs.CV)
Abstract
Articulated objects like cabinets and doors are widespread in daily life. However, directly manipulating 3D articulated objects is challenging because they have diverse geometrical shapes, semantic categories, and kinetic constraints. Prior works mostly focused on recognizing and manipulating articulated objects with specific joint types. They can either estimate the joint parameters or distinguish suitable grasp poses to facilitate trajectory planning. Although these approaches have succeeded in certain types of articulated objects, they lack generalizability to unseen objects, which significantly impedes their application in broader scenarios. In this paper, we propose a novel framework of Generalizable Articulation Modeling and Manipulating for Articulated Objects (GAMMA), which learns both articulation modeling and grasp pose affordance from diverse articulated objects with different categories. In addition, GAMMA adopts adaptive manipulation to iteratively reduce the modeling errors and enhance manipulation performance. We train GAMMA with the PartNet-Mobility dataset and evaluate with comprehensive experiments in SAPIEN simulation and real-world Franka robot arms. Results show that GAMMA significantly outperforms SOTA articulation modeling and manipulation algorithms in unseen and cross-category articulated objects. We will open-source all codes and datasets in both simulation and real robots for reproduction in the final version. Images and videos are published on the project website at: this http URL
TaxAI: A Dynamic Economic Simulator and Benchmark for Multi-Agent Reinforcement Learning
Authors: Qirui Mi, Siyu Xia, Yan Song, Haifeng Zhang, Shenghao Zhu, Jun Wang
Subjects: Computational Engineering, Finance, and Science (cs.CE)
Abstract
Taxation and government spending are crucial tools for governments to promote economic growth and maintain social equity. However, the difficulty in accurately predicting the dynamic strategies of diverse self-interested households presents a challenge for governments to implement effective tax policies. Given its proficiency in modeling other agents in partially observable environments and adaptively learning to find optimal policies, Multi-Agent Reinforcement Learning (MARL) is highly suitable for solving dynamic games between the government and numerous households. Although MARL shows more potential than traditional methods such as the genetic algorithm and dynamic programming, there is a lack of large-scale multi-agent reinforcement learning economic simulators. Therefore, we propose a MARL environment, named \textbf{TaxAI}, for dynamic games involving $N$ households, government, firms, and financial intermediaries based on the Bewley-Aiyagari economic model. Our study benchmarks 2 traditional economic methods with 7 MARL methods on TaxAI, demonstrating the effectiveness and superiority of MARL algorithms. Moreover, TaxAI's scalability in simulating dynamic interactions between the government and 10,000 households, coupled with real-data calibration, grants it a substantial improvement in scale and reality over existing simulators. Therefore, TaxAI is the most realistic economic simulator, which aims to generate feasible recommendations for governments and individuals.
Meshless interface tracking for the simulation of dendrite envelope growth
Authors: Mitja Jančič, Miha Založnik, Gregor Kosec
Abstract
The growth of dendritic grains during solidification is often modelled using the Grain Envelope Model (GEM), in which the envelope of the dendrite is an interface tracked by the Phase Field Interface Capturing (PFIC) method. In the PFIC method, an phase-field equation is solved on a fixed mesh to track the position of the envelope. While being versatile and robust, PFIC introduces certain numerical artefacts. In this work, we present an alternative approach for the solution of the GEM that employs a Meshless (sharp) Interface Tracking (MIT) formulation, which uses direct, artefact-free interface tracking. In the MIT, the envelope (interface) is defined as a moving domain boundary and the interface-tracking nodes are boundary nodes for the diffusion problem solved in the domain. To increase the accuracy of the method for the diffusion-controlled moving-boundary problem, an \h-adaptive spatial discretization is used, thus, the node spacing is refined in the vicinity of the envelope. MIT combines a parametric surface reconstruction, a mesh-free discretization of the parametric surfaces and the space enclosed by them, and a high-order approximation of the partial differential operators and of the solute concentration field using radial basis functions augmented with monomials. The proposed method is demonstrated on a two-dimensional \h-adaptive solution of the diffusive growth of dendrite and evaluated by comparing the results to the PFIC approach. It is shown that MIT can reproduce the results calculated with PFIC, that it is convergent and that it can capture more details in the envelope shape than PFIC with a similar spatial discretization.
HTC-DC Net: Monocular Height Estimation from Single Remote Sensing Images
Authors: Sining Chen, Yilei Shi, Zhitong Xiong, Xiao Xiang Zhu
Subjects: Computer Vision and Pattern Recognition (cs.CV)
Abstract
3D geo-information is of great significance for understanding the living environment; however, 3D perception from remote sensing data, especially on a large scale, is restricted. To tackle this problem, we propose a method for monocular height estimation from optical imagery, which is currently one of the richest sources of remote sensing data. As an ill-posed problem, monocular height estimation requires well-designed networks for enhanced representations to improve performance. Moreover, the distribution of height values is long-tailed with the low-height pixels, e.g., the background, as the head, and thus trained networks are usually biased and tend to underestimate building heights. To solve the problems, instead of formalizing the problem as a regression task, we propose HTC-DC Net following the classification-regression paradigm, with the head-tail cut (HTC) and the distribution-based constraints (DCs) as the main contributions. HTC-DC Net is composed of the backbone network as the feature extractor, the HTC-AdaBins module, and the hybrid regression process. The HTC-AdaBins module serves as the classification phase to determine bins adaptive to each input image. It is equipped with a vision transformer encoder to incorporate local context with holistic information and involves an HTC to address the long-tailed problem in monocular height estimation for balancing the performances of foreground and background pixels. The hybrid regression process does the regression via the smoothing of bins from the classification phase, which is trained via DCs. The proposed network is tested on three datasets of different resolutions, namely ISPRS Vaihingen (0.09 m), DFC19 (1.3 m) and GBH (3 m). Experimental results show the superiority of the proposed network over existing methods by large margins. Extensive ablation studies demonstrate the effectiveness of each design component.
Accurate and lightweight dehazing via multi-receptive-field non-local network and novel contrastive regularization
Abstract
Recently, deep learning-based methods have dominated image dehazing domain. Although very competitive dehazing performance has been achieved with sophisticated models, effective solutions for extracting useful features are still under-explored. In addition, non-local network, which has made a breakthrough in many vision tasks, has not been appropriately applied to image dehazing. Thus, a multi-receptive-field non-local network (MRFNLN) consisting of the multi-stream feature attention block (MSFAB) and cross non-local block (CNLB) is presented in this paper. We start with extracting richer features for dehazing. Specifically, we design a multi-stream feature extraction (MSFE) sub-block, which contains three parallel convolutions with different receptive fields (i.e., $1\times 1$, $3\times 3$, $5\times 5$) for extracting multi-scale features. Following MSFE, we employ an attention sub-block to make the model adaptively focus on important channels/regions. The MSFE and attention sub-blocks constitute our MSFAB. Then, we design a cross non-local block (CNLB), which can capture long-range dependencies beyond the query. Instead of the same input source of query branch, the key and value branches are enhanced by fusing more preceding features. CNLB is computation-friendly by leveraging a spatial pyramid down-sampling (SPDS) strategy to reduce the computation and memory consumption without sacrificing the performance. Last but not least, a novel detail-focused contrastive regularization (DFCR) is presented by emphasizing the low-level details and ignoring the high-level semantic information in the representation space. Comprehensive experimental results demonstrate that the proposed MRFNLN model outperforms recent state-of-the-art dehazing methods with less than 1.5 Million parameters.
Adaptive Output-Feedback Model Predictive Control of Hammerstein Systems with Unknown Linear Dynamics
Authors: Mohammadreza Kamaldar, Dennis S. Bernstein
Subjects: Systems and Control (eess.SY); Dynamical Systems (math.DS); Optimization and Control (math.OC)
Abstract
This paper considers model predictive control of Hammerstein systems, where the linear dynamics are a priori unknown and the input nonlinearity is known. Predictive cost adaptive control (PCAC) is applied to this system using recursive least squares for online, closed-loop system identification with optimization over a receding horizon performed by quadratic programming (QP). In order to account for the input nonlinearity, the input matrix is defined to be control dependent, and the optimization is performed iteratively. This technique is applied to output stabilization of a chain of integrators with unknown dynamics under control saturation and deadzone input nonlinearity.
FLIP: Cross-domain Face Anti-spoofing with Language Guidance
Abstract
Face anti-spoofing (FAS) or presentation attack detection is an essential component of face recognition systems deployed in security-critical applications. Existing FAS methods have poor generalizability to unseen spoof types, camera sensors, and environmental conditions. Recently, vision transformer (ViT) models have been shown to be effective for the FAS task due to their ability to capture long-range dependencies among image patches. However, adaptive modules or auxiliary loss functions are often required to adapt pre-trained ViT weights learned on large-scale datasets such as ImageNet. In this work, we first show that initializing ViTs with multimodal (e.g., CLIP) pre-trained weights improves generalizability for the FAS task, which is in line with the zero-shot transfer capabilities of vision-language pre-trained (VLP) models. We then propose a novel approach for robust cross-domain FAS by grounding visual representations with the help of natural language. Specifically, we show that aligning the image representation with an ensemble of class descriptions (based on natural language semantics) improves FAS generalizability in low-data regimes. Finally, we propose a multimodal contrastive learning strategy to boost feature generalization further and bridge the gap between source and target domains. Extensive experiments on three standard protocols demonstrate that our method significantly outperforms the state-of-the-art methods, achieving better zero-shot transfer performance than five-shot transfer of adaptive ViTs. Code: https://github.com/koushiksrivats/FLIP
SA2-Net: Scale-aware Attention Network for Microscopic Image Segmentation
Authors: Mustansar Fiaz, Moein Heidari, Rao Muhammad Anwer, Hisham Cholakkal
Subjects: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
Abstract
Microscopic image segmentation is a challenging task, wherein the objective is to assign semantic labels to each pixel in a given microscopic image. While convolutional neural networks (CNNs) form the foundation of many existing frameworks, they often struggle to explicitly capture long-range dependencies. Although transformers were initially devised to address this issue using self-attention, it has been proven that both local and global features are crucial for addressing diverse challenges in microscopic images, including variations in shape, size, appearance, and target region density. In this paper, we introduce SA2-Net, an attention-guided method that leverages multi-scale feature learning to effectively handle diverse structures within microscopic images. Specifically, we propose scale-aware attention (SA2) module designed to capture inherent variations in scales and shapes of microscopic regions, such as cells, for accurate segmentation. This module incorporates local attention at each level of multi-stage features, as well as global attention across multiple resolutions. Furthermore, we address the issue of blurred region boundaries (e.g., cell boundaries) by introducing a novel upsampling strategy called the Adaptive Up-Attention (AuA) module. This module enhances the discriminative ability for improved localization of microscopic regions using an explicit attention mechanism. Extensive experiments on five challenging datasets demonstrate the benefits of our SA2-Net model. Our source code is publicly available at \url{https://github.com/mustansarfiaz/SA2-Net}.
Keyword: quantization
Highly Efficient SNNs for High-speed Object Detection
Abstract
The high biological properties and low energy consumption of Spiking Neural Networks (SNNs) have brought much attention in recent years. However, the converted SNNs generally need large time steps to achieve satisfactory performance, which will result in high inference latency and computational resources increase. In this work, we propose a highly efficient and fast SNN for object detection. First, we build an initial compact ANN by using quantization training method of convolution layer fold batch normalization layer and neural network modification. Second, we theoretically analyze how to obtain the low complexity SNN correctly. Then, we propose a scale-aware pseudoquantization scheme to guarantee the correctness of the compact ANN to SNN. Third, we propose a continuous inference scheme by using a Feed-Forward Integrate-and-Fire (FewdIF) neuron to realize high-speed object detection. Experimental results show that our efficient SNN can achieve 118X speedup on GPU with only 1.5MB parameters for object detection tasks. We further verify our SNN on FPGA platform and the proposed model can achieve 800+FPS object detection with extremely low latency.
ModuLoRA: Finetuning 3-Bit LLMs on Consumer GPUs by Integrating with Modular Quantizers
Authors: Junjie Yin, Jiahao Dong, Yingheng Wang, Christopher De Sa, Volodymyr Kuleshov
Abstract
We propose a memory-efficient finetuning algorithm for large language models (LLMs) that supports finetuning LLMs with 65B parameters in 3-bit or 4-bit precision on as little as one 48GB GPU. Our method, modular low-rank adaptation (ModuLoRA), integrates any user-specified weight quantizer with finetuning via low-rank adapters (LoRAs). Our approach relies on a simple quantization-agnostic backward pass that adaptively materializes low-precision LLM weights from a custom black-box quantization module. This approach enables finetuning 3-bit LLMs for the first time--leveraging state-of-the-art 3-bit OPTQ quantization often outperforms finetuning that relies on less sophisticated 4-bit and 8-bit methods. In our experiments, ModuLoRA attains competitive performance on text classification, natural language infernece, and instruction following tasks using significantly less memory than existing approaches, and we also surpass the state-of-the-art ROUGE score on a popular summarization task. We release ModuLoRA together with a series of low-precision models--including the first family of 3-bit instruction following Alpaca LLMs--as part of LLMTOOLS, a user-friendly library for quantizing, running, and finetuning LLMs on consumer GPUs.
Keyword: efficient
Highly Efficient SNNs for High-speed Object Detection
Zero-Shot and Few-Shot Video Question Answering with Multi-Modal Prompts
Towards Efficient and Trustworthy AI Through Hardware-Algorithm-Communication Co-Design
Scalable Multi-Robot Collaboration with Large Language Models: Centralized or Decentralized Systems?
Accelerating Motion Planning via Optimal Transport
Loupe: Driving the Development of OS Compatibility Layers
Robust Blockwise Random Pivoting: Fast and Accurate Adaptive Interpolative Decomposition
GNNHLS: Evaluating Graph Neural Network Inference via High-Level Synthesis
Symbolic Imitation Learning: From Black-Box to Explainable Driving Policies
Bridging the complexity gap in Tbps-achieving THz-band baseband processing
Channel Estimation for Reconfigurable Intelligent Surface-Aided Multiuser Communication Systems Exploiting Statistical CSI of Correlated RIS-User Channels
Effective Long-Context Scaling of Foundation Models
A Multiscale Finite Element Method for an Elliptic Distributed Optimal Control Problem with Rough Coefficients and Control Constraints
Label Augmentation Method for Medical Landmark Detection in Hip Radiograph Images
A Modular Bio-inspired Robotic Hand with High Sensitivity
A priori and a posteriori error analysis for a VEM discretization of the convection-diffusion eigenvalue problem
Differentiable Robot Neural Distance Function for Adaptive Grasp Synthesis on a Unified Robotic Arm-Hand System
Discovering Utility-driven Interval Rules
Learning Effective NeRFs and SDFs Representations with 3D Generative Adversarial Networks for 3D Object Generation: Technical Report for ICCV 2023 OmniObject3D Challenge
E2Net: Resource-Efficient Continual Learning with Elastic Expansion Network
ModuLoRA: Finetuning 3-Bit LLMs on Consumer GPUs by Integrating with Modular Quantizers
UVL: A Unified Framework for Video Tampering Localization
Generative Semi-supervised Learning with Meta-Optimized Synthetic Samples
The Trickle-down Impact of Reward (In-)consistency on RLHF
Doppler Time-of-Flight Rendering
Learning to Terminate in Object Navigation
Max-Sliced Mutual Information
Hybrid Digital-Wave Domain Channel Estimator for Stacked Intelligent Metasurface Enabled Multi-User MISO Systems
DiffGAN-F2S: Symmetric and Efficient Denoising Diffusion GANs for Structural Connectivity Prediction from Brain fMRI
Parameter-Saving Adversarial Training: Reinforcing Multi-Perturbation Robustness via Hypernetworks
Controllable Text Generation with Residual Memory Transformer
Beyond Reverse KL: Generalizing Direct Preference Optimization with Diverse Divergence Constraints
Nondestructive chicken egg fertility detection using CNN-transfer learning algorithms
Efficiency Separation between RL Methods: Model-Free, Model-Based and Goal-Conditioned
Multi-scale Recurrent LSTM and Transformer Network for Depth Completion
LagrangeBench: A Lagrangian Fluid Mechanics Benchmarking Suite
Transformer-VQ: Linear-Time Transformers via Vector Quantization
Nek5000/RS Performance on Advanced GPU Architectures
Genetic Engineering Algorithm (GEA): An Efficient Metaheuristic Algorithm for Solving Combinatorial Optimization Problems
Nonlinear MPC design for incrementally ISS systems with application to GRU networks
A Small World of Bad Guys: Investigating the Behavior of Hacker Groups in Cyber-Attacks
Hadoop-Oriented SVM-LRU (H-SVM-LRU): An Intelligent Cache Replacement Algorithm to Improve MapReduce Performance
Efficient Hardware Implementation of Constant Time Sampling for HQC
Capturing Requirements for a Data Annotation Tool for Intensive Care: Experimental User-Centered Design Study
SwiftSpatial: Spatial Joins on Modern Hardware
HOI4ABOT: Human-Object Interaction Anticipation for Human Intention Reading Collaborative roBOTs
Audio-Visual Speaker Verification via Joint Cross-Attention
Compilation as a Defense: Enhancing DL Model Attack Robustness via Tensor Optimization
Deep Learning Based Uplink Multi-User SIMO Beamforming Design
Robust Offline Reinforcement Learning -- Certify the Confidence Interval
Sparse Submodular Function Minimization
ConceptGraphs: Open-Vocabulary 3D Scene Graphs for Perception and Planning
DreamGaussian: Generative Gaussian Splatting for Efficient 3D Content Creation
Training a Large Video Model on a Single Machine in a Day
Keyword: faster
CLIP-Hand3D: Exploiting 3D Hand Pose Estimation via Context-Aware Prompting
Distill to Delete: Unlearning in Graph Networks with Knowledge Distillation
Matrix Multiplication Verification Using Coding Theory
Hierarchical Network Data Analytics Framework for B5G Network Automation: Design and Implementation
DeepPCR: Parallelizing Sequential Operations in Neural Networks
Efficient Supervised Training of Audio Transformers for Music Representation Learning
HOI4ABOT: Human-Object Interaction Anticipation for Human Intention Reading Collaborative roBOTs
Perceiving Extrinsic Contacts from Touch Improves Learning Insertion Policies
DreamGaussian: Generative Gaussian Splatting for Efficient 3D Content Creation
Visual In-Context Learning for Few-Shot Eczema Segmentation
HyperPPO: A scalable method for finding small policies for robotic control
Keyword: mobile
Breamy: An augmented reality mHealth prototype for surgical decision-making in breast cancer
Dynamics of Ideological Biases of Social Media Users
DynaCon: Dynamic Robot Planner with Contextual Awareness via LLMs
Nondestructive chicken egg fertility detection using CNN-transfer learning algorithms
Semantic Scene Difference Detection in Daily Life Patroling by Mobile Robots using Pre-Trained Large-Scale Vision-Language Model
Review of Machine Learning Methods for Additive Manufacturing of Functionally Graded Materials
Keyword: pruning
Discovering Utility-driven Interval Rules
Rethinking Domain Generalization: Discriminability and Generalizability
DreamGaussian: Generative Gaussian Splatting for Efficient 3D Content Creation
Keyword: diffusion
A priori and a posteriori error analysis for a VEM discretization of the convection-diffusion eigenvalue problem
Compositional Sculpting of Iterative Generative Processes
DiffGAN-F2S: Symmetric and Efficient Denoising Diffusion GANs for Structural Connectivity Prediction from Brain fMRI
Object Motion Guided Human Motion Synthesis
DeepPCR: Parallelizing Sequential Operations in Neural Networks
Meshless interface tracking for the simulation of dendrite envelope growth
Distilling ODE Solvers of Diffusion Models into Smaller Steps
KV Inversion: KV Embeddings Learning for Text-Conditioned Real Image Action Editing
Keyword: adaptive
Enhancing Cross-Category Learning in Recommendation Systems with Multi-Layer Embedding Training
Improving Adaptive Online Learning Using Refined Discretization
A review of variable-pitch propellers and their control strategies in aerospace systems
Differentiable Robot Neural Distance Function for Adaptive Grasp Synthesis on a Unified Robotic Arm-Hand System
ModuLoRA: Finetuning 3-Bit LLMs on Consumer GPUs by Integrating with Modular Quantizers
Joint Correcting and Refinement for Balanced Low-Light Image Enhancement
Context-I2W: Mapping Images to Context-dependent Words for Accurate Zero-Shot Composed Image Retrieval
Adaptive Real-Time Numerical Differentiation with Variable-Rate Forgetting and Exponential Resetting
Laboratory Automation: Precision Insertion with Adaptive Fingers utilizing Contact through Sliding with Tactile-based Pose Estimation
Cloth2Body: Generating 3D Human Body Mesh from 2D Clothing
GAMMA: Generalizable Articulation Modeling and Manipulation for Articulated Objects
TaxAI: A Dynamic Economic Simulator and Benchmark for Multi-Agent Reinforcement Learning
Meshless interface tracking for the simulation of dendrite envelope growth
HTC-DC Net: Monocular Height Estimation from Single Remote Sensing Images
Accurate and lightweight dehazing via multi-receptive-field non-local network and novel contrastive regularization
Adaptive Output-Feedback Model Predictive Control of Hammerstein Systems with Unknown Linear Dynamics
FLIP: Cross-domain Face Anti-spoofing with Language Guidance
SA2-Net: Scale-aware Attention Network for Microscopic Image Segmentation
Keyword: quantization
Highly Efficient SNNs for High-speed Object Detection
ModuLoRA: Finetuning 3-Bit LLMs on Consumer GPUs by Integrating with Modular Quantizers