Abstract
Quantization (Alistarh et al., 2017) is an important (stochastic) compression technique that reduces the volume of transmitted bits during each communication round in distributed model training. Suresh et al. (2022) introduce correlated quantizers and show their advantages over independent counterparts by analyzing distributed SGD communication complexity. We analyze the forefront distributed non-convex optimization algorithm MARINA (Gorbunov et al., 2022) utilizing the proposed correlated quantizers and show that it outperforms the original MARINA and distributed SGD of Suresh et al. (2022) with regard to the communication complexity. We significantly refine the original analysis of MARINA without any additional assumptions using the weighted Hessian variance (Tyurin et al., 2022), and then we expand the theoretical framework of MARINA to accommodate a substantially broader range of potentially correlated and biased compressors, thus dilating the applicability of the method beyond the conventional independent unbiased compressor setup. Extensive experimental results corroborate our theoretical findings.
Keyword: optimization
Dataset Optimization for Chronic Disease Prediction with Bio-Inspired Feature Selection
Authors: Authors: Abeer Dyoub, Ivan Letteri
Subjects: Neural and Evolutionary Computing (cs.NE); Machine Learning (cs.LG)
Abstract
In this study, we investigated the application of bio-inspired optimization algorithms, including Genetic Algorithm, Particle Swarm Optimization, and Whale Optimization Algorithm, for feature selection in chronic disease prediction. The primary goal was to enhance the predictive accuracy of models streamline data dimensionality, and make predictions more interpretable and actionable. The research encompassed a comparative analysis of the three bio-inspired feature selection approaches across diverse chronic diseases, including diabetes, cancer, kidney, and cardiovascular diseases. Performance metrics such as accuracy, precision, recall, and f1 score are used to assess the effectiveness of the algorithms in reducing the number of features needed for accurate classification. The results in general demonstrate that the bio-inspired optimization algorithms are effective in reducing the number of features required for accurate classification. However, there have been variations in the performance of the algorithms on different datasets. The study highlights the importance of data pre-processing and cleaning in ensuring the reliability and effectiveness of the analysis. This study contributes to the advancement of predictive analytics in the realm of chronic diseases. The potential impact of this work extends to early intervention, precision medicine, and improved patient outcomes, providing new avenues for the delivery of healthcare services tailored to individual needs. The findings underscore the potential benefits of using bio-inspired optimization algorithms for feature selection in chronic disease prediction, offering valuable insights for improving healthcare outcomes.
Abstract
While machine learning models are typically trained to solve prediction problems, we might often want to use them for optimization problems. For example, given a dataset of proteins and their corresponding fluorescence levels, we might want to optimize for a new protein with the highest possible fluorescence. This kind of data-driven optimization (DDO) presents a range of challenges beyond those in standard prediction problems, since we need models that successfully predict the performance of new designs that are better than the best designs seen in the training set. It is not clear theoretically when existing approaches can even perform better than the naive approach that simply selects the best design in the dataset. In this paper, we study how structure can enable sample-efficient data-driven optimization. To formalize the notion of structure, we introduce functional graphical models (FGMs) and show theoretically how they can provide for principled data-driven optimization by decomposing the original high-dimensional optimization problem into smaller sub-problems. This allows us to derive much more practical regret bounds for DDO, and the result implies that DDO with FGMs can achieve nearly optimal designs in situations where naive approaches fail due to insufficient coverage of the offline data. We further present a data-driven optimization algorithm that inferes the FGM structure itself, either over the original input variables or a latent variable representation of the inputs.
Standardizing Your Training Process for Human Activity Recognition Models: A Comprehensive Review in the Tunable Factors
Authors: Authors: Yiran Huang, Haibin Zhao, Yexu Zhou, Till Riedel, Michael Beigl
Abstract
In recent years, deep learning has emerged as a potent tool across a multitude of domains, leading to a surge in research pertaining to its application in the wearable human activity recognition (WHAR) domain. Despite the rapid development, concerns have been raised about the lack of standardization and consistency in the procedures used for experimental model training, which may affect the reproducibility and reliability of research results. In this paper, we provide an exhaustive review of contemporary deep learning research in the field of WHAR and collate information pertaining to the training procedure employed in various studies. Our findings suggest that a major trend is the lack of detail provided by model training protocols. Besides, to gain a clearer understanding of the impact of missing descriptions, we utilize a control variables approach to assess the impact of key tunable components (e.g., optimization techniques and early stopping criteria) on the inter-subject generalization capabilities of HAR models. With insights from the analyses, we define a novel integrated training procedure tailored to the WHAR model. Empirical results derived using five well-known \ac{whar} benchmark datasets and three classical HAR model architectures demonstrate the effectiveness of our proposed methodology: in particular, there is a significant improvement in macro F1 leave one subject out cross-validation performance.
Optimized Ensemble Model Towards Secured Industrial IoT Devices
Authors: Authors: MohammadNoor Injadat
Subjects: Cryptography and Security (cs.CR); Artificial Intelligence (cs.AI)
Abstract
The continued growth in the deployment of Internet-of-Things (IoT) devices has been fueled by the increased connectivity demand, particularly in industrial environments. However, this has led to an increase in the number of network related attacks due to the increased number of potential attack surfaces. Industrial IoT (IIoT) devices are prone to various network related attacks that can have severe consequences on the manufacturing process as well as on the safety of the workers in the manufacturing plant. One promising solution that has emerged in recent years for attack detection is Machine learning (ML). More specifically, ensemble learning models have shown great promise in improving the performance of the underlying ML models. Accordingly, this paper proposes a framework based on the combined use of Bayesian Optimization-Gaussian Process (BO-GP) with an ensemble tree-based learning model to improve the performance of intrusion and attack detection in IIoT environments. The proposed framework's performance is evaluated using the Windows 10 dataset collected by the Cyber Range and IoT labs at University of New South Wales. Experimental results illustrate the improvement in detection accuracy, precision, and F-score when compared to standard tree and ensemble tree models.
FPRF: Feed-Forward Photorealistic Style Transfer of Large-Scale 3D Neural Radiance Fields
Authors: Authors: GeonU Kim, Kim Youwang, Tae-Hyun Oh
Abstract
We present FPRF, a feed-forward photorealistic style transfer method for large-scale 3D neural radiance fields. FPRF stylizes large-scale 3D scenes with arbitrary, multiple style reference images without additional optimization while preserving multi-view appearance consistency. Prior arts required tedious per-style/-scene optimization and were limited to small-scale 3D scenes. FPRF efficiently stylizes large-scale 3D scenes by introducing a style-decomposed 3D neural radiance field, which inherits AdaIN's feed-forward stylization machinery, supporting arbitrary style reference images. Furthermore, FPRF supports multi-reference stylization with the semantic correspondence matching and local AdaIN, which adds diverse user control for 3D scene styles. FPRF also preserves multi-view consistency by applying semantic matching and style transfer processes directly onto queried features in 3D space. In experiments, we demonstrate that FPRF achieves favorable photorealistic quality 3D scene stylization for large-scale scenes with diverse reference images. Project page: https://kim-geonu.github.io/FPRF/
Correlated Quantization for Faster Nonconvex Distributed Optimization
Authors: Authors: Andrei Panferov, Yury Demidovich, Ahmad Rammal, Peter Richtárik
Subjects: Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Distributed, Parallel, and Cluster Computing (cs.DC); Optimization and Control (math.OC)
Abstract
Quantization (Alistarh et al., 2017) is an important (stochastic) compression technique that reduces the volume of transmitted bits during each communication round in distributed model training. Suresh et al. (2022) introduce correlated quantizers and show their advantages over independent counterparts by analyzing distributed SGD communication complexity. We analyze the forefront distributed non-convex optimization algorithm MARINA (Gorbunov et al., 2022) utilizing the proposed correlated quantizers and show that it outperforms the original MARINA and distributed SGD of Suresh et al. (2022) with regard to the communication complexity. We significantly refine the original analysis of MARINA without any additional assumptions using the weighted Hessian variance (Tyurin et al., 2022), and then we expand the theoretical framework of MARINA to accommodate a substantially broader range of potentially correlated and biased compressors, thus dilating the applicability of the method beyond the conventional independent unbiased compressor setup. Extensive experimental results corroborate our theoretical findings.
Towards Safe Load Balancing based on Control Barrier Functions and Deep Reinforcement Learning
Authors: Authors: Lam Dinh, Pham Tran Anh Quang, Jérémie Leguay
Subjects: Networking and Internet Architecture (cs.NI); Machine Learning (cs.LG)
Abstract
Deep Reinforcement Learning (DRL) algorithms have recently made significant strides in improving network performance. Nonetheless, their practical use is still limited in the absence of safe exploration and safe decision-making. In the context of commercial solutions, reliable and safe-to-operate systems are of paramount importance. Taking this problem into account, we propose a safe learning-based load balancing algorithm for Software Defined-Wide Area Network (SD-WAN), which is empowered by Deep Reinforcement Learning (DRL) combined with a Control Barrier Function (CBF). It safely projects unsafe actions into feasible ones during both training and testing, and it guides learning towards safe policies. We successfully implemented the solution on GPU to accelerate training by approximately 110x times and achieve model updates for on-policy methods within a few seconds, making the solution practical. We show that our approach delivers near-optimal Quality-of-Service (QoS performance in terms of end-to-end delay while respecting safety requirements related to link capacity constraints. We also demonstrated that on-policy learning based on Proximal Policy Optimization (PPO) performs better than off-policy learning with Deep Deterministic Policy Gradient (DDPG) when both are combined with a CBF for safe load balancing.
Multi-objective Feature Selection in Remote Health Monitoring Applications
Authors: Authors: Le Ngu Nguyen, Constantino Álvarez Casado, Manuel Lage Cañellas, Anirban Mukherjee, Nhi Nguyen, Dinesh Babu Jayagopi, Miguel Bordallo López
Abstract
Radio frequency (RF) signals have facilitated the development of non-contact human monitoring tasks, such as vital signs measurement, activity recognition, and user identification. In some specific scenarios, an RF signal analysis framework may prioritize the performance of one task over that of others. In response to this requirement, we employ a multi-objective optimization approach inspired by biological principles to select discriminative features that enhance the accuracy of breathing patterns recognition while simultaneously impeding the identification of individual users. This approach is validated using a novel vital signs dataset consisting of 50 subjects engaged in four distinct breathing patterns. Our findings indicate a remarkable result: a substantial divergence in accuracy between breathing recognition and user identification. As a complementary viewpoint, we present a contrariwise result to maximize user identification accuracy and minimize the system's capacity for breathing activity recognition.
Optimization by linear kinetic equations and mean-field Langevin dynamics
Abstract
Probably one of the most striking examples of the close connections between global optimization processes and statistical physics is the simulated annealing method, inspired by the famous Monte Carlo algorithm devised by Metropolis et al. in the middle of the last century. In this paper we show how the tools of linear kinetic theory allow to describe this gradient-free algorithm from the perspective of statistical physics and how convergence to the global minimum can be related to classical entropy inequalities. This analysis highlight the strong link between linear Boltzmann equations and stochastic optimization methods governed by Markov processes. Thanks to this formalism we can establish the connections between the simulated annealing process and the corresponding mean-field Langevin dynamics characterized by a stochastic gradient descent approach. Generalizations to other selection strategies in simulated annealing that avoid the acceptance-rejection dynamic are also provided.
Characterising the take-off dynamics and energy efficiency in spring-driven jumping robots
Abstract
Previous design methodologies for spring-driven jumping robots focused on jump height optimization for specific tasks. In doing so, numerous designs have been proposed including using nonlinear spring-linkages to increase the elastic energy storage and jump height. However, these systems can never achieve their theoretical maximum jump height due to taking off before the spring energy is fully released, resulting in an incomplete transfer of stored elastic energy to gravitational potential energy. This paper presents low-order models aimed at characterising the energy conversion during the acceleration phase of jumping. It also proposes practical solutions for increasing the energy efficiency of jumping robots. A dynamic analysis is conducted on a multibody system comprised of rotational links, which is experimentally validated using a physical demonstrator. The analysis reveals that inefficient energy conversion is attributed to inertial effects caused by rotational and unsprung masses. Since these masses cannot be entirely eliminated from a physical linkage, a practical approach to improving energy efficiency involves structural redesign to reduce structural mass and moments of inertia while maintaining compliance with structural strength and stiffness requirements.
Graph Q-Learning for Combinatorial Optimization
Authors: Authors: Victoria M. Dax, Jiachen Li, Kevin Leahy, Mykel J. Kochenderfer
Abstract
Graph-structured data is ubiquitous throughout natural and social sciences, and Graph Neural Networks (GNNs) have recently been shown to be effective at solving prediction and inference problems on graph data. In this paper, we propose and demonstrate that GNNs can be applied to solve Combinatorial Optimization (CO) problems. CO concerns optimizing a function over a discrete solution space that is often intractably large. To learn to solve CO problems, we formulate the optimization process as a sequential decision making problem, where the return is related to how close the candidate solution is to optimality. We use a GNN to learn a policy to iteratively build increasingly promising candidate solutions. We present preliminary evidence that GNNs trained through Q-Learning can solve CO problems with performance approaching state-of-the-art heuristic-based solvers, using only a fraction of the parameters and training time.
MatSAM: Efficient Materials Microstructure Extraction via Visual Large Model
Authors: Authors: Changtai Li, Xu Han, Chao Yao, Xiaojuan Ban
Subjects: Computer Vision and Pattern Recognition (cs.CV)
Abstract
Accurate and efficient extraction of microstructures in microscopic images of materials plays a critical role in the exploration of structure-property relationships and the optimization of process parameters. Deep learning-based image segmentation techniques that rely on manual annotation are time-consuming and labor-intensive and hardly meet the demand for model transferability and generalization. Segment Anything Model (SAM), a large visual model with powerful deep feature representation and zero-shot generalization capabilities, has provided new solutions for image segmentation. However, directly applying SAM to segmenting microstructures in microscopic images of materials without human annotation cannot achieve the expected results, as the difficulty of adapting its native prompt engineering to the dense and dispersed characteristics of key microstructures in materials microscopy images. In this paper, we propose MatSAM, a general and efficient microstructure extraction solution based on SAM. A new point-based prompts generation strategy is designed, grounded on the distribution and shape of materials microstructures. It generates prompts for different microscopic images, fuses the prompts of the region of interest (ROI) key points and grid key points, and integrates post-processing methods for quantitative characterization of materials microstructures. For common microstructures including grain boundary and phase, MatSAM achieves superior segmentation performance to conventional methods and is even preferable to supervised learning methods evaluated on 18 materials microstructures imaged by the optical microscope (OM) and scanning electron microscope (SEM). We believe that MatSAM can significantly reduce the cost of quantitative characterization of materials microstructures and accelerate the design of new materials.
Parrot: Pareto-optimal Multi-Reward Reinforcement Learning Framework for Text-to-Image Generation
Authors: Authors: Seung Hyun Lee, Yinxiao Li, Junjie Ke, Innfarn Yoo, Han Zhang, Jiahui Yu, Qifei Wang, Fei Deng, Glenn Entis, Junfeng He, Gang Li, Sangpil Kim, Irfan Essa, Feng Yang
Subjects: Computer Vision and Pattern Recognition (cs.CV)
Abstract
Recent works demonstrate that using reinforcement learning (RL) with quality rewards can enhance the quality of generated images in text-to-image (T2I) generation. However, a simple aggregation of multiple rewards may cause over-optimization in certain metrics and degradation in others, and it is challenging to manually find the optimal weights. An effective strategy to jointly optimize multiple rewards in RL for T2I generation is highly desirable. This paper introduces Parrot, a novel multi-reward RL framework for T2I generation. Through the use of the batch-wise Pareto optimal selection, Parrot automatically identifies the optimal trade-off among different rewards during the RL optimization of the T2I generation. Additionally, Parrot employs a joint optimization approach for the T2I model and the prompt expansion network, facilitating the generation of quality-aware text prompts, thus further enhancing the final image quality. To counteract the potential catastrophic forgetting of the original user prompt due to prompt expansion, we introduce original prompt centered guidance at inference time, ensuring that the generated image remains faithful to the user input. Extensive experiments and a user study demonstrate that Parrot outperforms several baseline methods across various quality criteria, including aesthetics, human preference, image sentiment, and text-image alignment.
Regularization of the discrete source problem in the nonlinear diffusive-logistic equation
Authors: Authors: Olga Krivorotko, Tatiana Zvonareva
Subjects: Numerical Analysis (math.NA); Optimization and Control (math.OC)
Abstract
A numerical algorithm for regularization of the solution of the source problem for the diffusion-logistic model based on information about the process at fixed moments of time of integral type has been developed. The peculiarity of the problem under study is the discrete formulation in space and impossibility to apply classical algorithms for its numerical solution. The regularization of the problem is based on the application of A.N. Tikhonov's approach and a priori information about the source of the process. The problem was formulated in a variational formulation and solved by the global tensor optimization method. It is shown that in the case of noisy data regularization improves the accuracy of the reconstructed source.
CAT-LLM: Prompting Large Language Models with Text Style Definition for Chinese Article-style Transfer
Abstract
Text style transfer is increasingly prominent in online entertainment and social media. However, existing research mainly concentrates on style transfer within individual English sentences, while ignoring the complexity of long Chinese texts, which limits the wider applicability of style transfer in digital media realm. To bridge this gap, we propose a Chinese Article-style Transfer framework (CAT-LLM), leveraging the capabilities of Large Language Models (LLMs). CAT-LLM incorporates a bespoke, pluggable Text Style Definition (TSD) module aimed at comprehensively analyzing text features in articles, prompting LLMs to efficiently transfer Chinese article-style. The TSD module integrates a series of machine learning algorithms to analyze article-style from both words and sentences levels, thereby aiding LLMs thoroughly grasp the target style without compromising the integrity of the original text. In addition, this module supports dynamic expansion of internal style trees, showcasing robust compatibility and allowing flexible optimization in subsequent research. Moreover, we select five Chinese articles with distinct styles and create five parallel datasets using ChatGPT, enhancing the models' performance evaluation accuracy and establishing a novel paradigm for evaluating subsequent research on article-style transfer. Extensive experimental results affirm that CAT-LLM outperforms current research in terms of transfer accuracy and content preservation, and has remarkable applicability to various types of LLMs.
Kernelized Normalizing Constant Estimation: Bridging Bayesian Quadrature and Bayesian Optimization
Abstract
In this paper, we study the problem of estimating the normalizing constant $\int e^{-\lambda f(x)}dx$ through queries to the black-box function $f$, where $f$ belongs to a reproducing kernel Hilbert space (RKHS), and $\lambda$ is a problem parameter. We show that to estimate the normalizing constant within a small relative error, the level of difficulty depends on the value of $\lambda$: When $\lambda$ approaches zero, the problem is similar to Bayesian quadrature (BQ), while when $\lambda$ approaches infinity, the problem is similar to Bayesian optimization (BO). More generally, the problem varies between BQ and BO. We find that this pattern holds true even when the function evaluations are noisy, bringing new aspects to this topic. Our findings are supported by both algorithm-independent lower bounds and algorithmic upper bounds, as well as simulation studies conducted on a variety of benchmark functions.
STAR-RIS Enhanced UAV-Enabled MEC Networks with Bi-Directional Task Offloading
Authors: Authors: Han Xiao, Xiaoyan Hu, Pengcheng Mu, Weile Zhang, Wenjie Wang, Kai-Kit Wong, Kun Yang
Subjects: Information Theory (cs.IT); Signal Processing (eess.SP)
Abstract
A simultaneously transmitting and reflecting reconfigurable intelligent surface (STAR-RIS) enhanced unnamed aerial vehicle (UAV)-enabled multi-user mobile edge computing (MEC) scheme is proposed in this paper. Different from the existing MEC works, the proposed scheme allows bi-directional offloading where users can simultaneously offload their computing tasks to the MEC servers situated at the ground base station (BS) and aerial UAV with the assistance of the STARRIS. Specifically, we formulate an optimization problem aiming at maximizing the energy efficiency of the system while ensuring the quality of service (QoS) constraint by jointly optimizing the resource allocation, user scheduling, passive beamforming of the STAR-RIS, and the UAV trajectory. An iterative algorithm designed with the Dinkelbach's algorithm and the successive convex approximation (SCA) is proposed to effectively handle the formulated non-convex optimization problem. Simulation results indicate that the proposed STAR-RIS enhanced UAV-enabled MEC scheme possesses significant advantages in enhancing the system energy efficiency over other baseline schemes including the conventional RIS-aided scheme.
An experimental evaluation of Deep Reinforcement Learning algorithms for HVAC control
Authors: Authors: Antonio Manjavacas, Alejandro Campoy-Nieves, Javier Jiménez-Raboso, Miguel Molina-Solana, Juan Gómez-Romero
Subjects: Machine Learning (cs.LG); Systems and Control (eess.SY)
Abstract
Heating, Ventilation, and Air Conditioning (HVAC) systems are a major driver of energy consumption in commercial and residential buildings. Recent studies have shown that Deep Reinforcement Learning (DRL) algorithms can outperform traditional reactive controllers. However, DRL-based solutions are generally designed for ad hoc setups and lack standardization for comparison. To fill this gap, this paper provides a critical and reproducible evaluation, in terms of comfort and energy consumption, of several state-of-the-art DRL algorithms for HVAC control. The study examines the controllers' robustness, adaptability, and trade-off between optimization goals by using the Sinergym framework. The results obtained confirm the potential of DRL algorithms, such as SAC and TD3, in complex scenarios and reveal several challenges related to generalization and incremental learning.
Cross-Inlining Binary Function Similarity Detection
Authors: Authors: Ang Jia, Ming Fan, Xi Xu, Wuxia Jin, Haijun Wang, Ting Liu
Subjects: Software Engineering (cs.SE); Cryptography and Security (cs.CR)
Abstract
Binary function similarity detection plays an important role in a wide range of security applications. Existing works usually assume that the query function and target function share equal semantics and compare their full semantics to obtain the similarity. However, we find that the function mapping is more complex, especially when function inlining happens. In this paper, we will systematically investigate cross-inlining binary function similarity detection. We first construct a cross-inlining dataset by compiling 51 projects using 9 compilers, with 4 optimizations, to 6 architectures, with 2 inlining flags, which results in two datasets both with 216 combinations. Then we construct the cross-inlining function mappings by linking the common source functions in these two datasets. Through analysis of this dataset, we find that three cross-inlining patterns widely exist while existing work suffers when detecting cross-inlining binary function similarity. Next, we propose a pattern-based model named CI-Detector for cross-inlining matching. CI-Detector uses the attributed CFG to represent the semantics of binary functions and GNN to embed binary functions into vectors. CI-Detector respectively trains a model for these three cross-inlining patterns. Finally, the testing pairs are input to these three models and all the produced similarities are aggregated to produce the final similarity. We conduct several experiments to evaluate CI-Detector. Results show that CI-Detector can detect cross-inlining pairs with a precision of 81% and a recall of 97%, which exceeds all state-of-the-art works.
GO-NeRF: Generating Virtual Objects in Neural Radiance Fields
Abstract
Despite advances in 3D generation, the direct creation of 3D objects within an existing 3D scene represented as NeRF remains underexplored. This process requires not only high-quality 3D object generation but also seamless composition of the generated 3D content into the existing NeRF. To this end, we propose a new method, GO-NeRF, capable of utilizing scene context for high-quality and harmonious 3D object generation within an existing NeRF. Our method employs a compositional rendering formulation that allows the generated 3D objects to be seamlessly composited into the scene utilizing learned 3D-aware opacity maps without introducing unintended scene modification. Moreover, we also develop tailored optimization objectives and training strategies to enhance the model's ability to exploit scene context and mitigate artifacts, such as floaters, originating from 3D object generation within a scene. Extensive experiments on both feed-forward and $360^o$ scenes show the superior performance of our proposed GO-NeRF in generating objects harmoniously composited with surrounding scenes and synthesizing high-quality novel view images. Project page at {\url{https://daipengwa.github.io/GO-NeRF/}.
EraseDiff: Erasing Data Influence in Diffusion Models
Authors: Authors: Jing Wu, Trung Le, Munawar Hayat, Mehrtash Harandi
Subjects: Computer Vision and Pattern Recognition (cs.CV)
Abstract
In response to data protection regulations and the ``right to be forgotten'', in this work, we introduce an unlearning algorithm for diffusion models. Our algorithm equips a diffusion model with a mechanism to mitigate the concerns related to data memorization. To achieve this, we formulate the unlearning problem as a bi-level optimization problem, wherein the outer objective is to preserve the utility of the diffusion model on the remaining data. The inner objective aims to scrub the information associated with forgetting data by deviating the learnable generative process from the ground-truth denoising procedure. To solve the resulting bi-level problem, we adopt a first-order method, having superior practical performance while being vigilant about the diffusion process and solving a bi-level problem therein. Empirically, we demonstrate that our algorithm can preserve the model utility, effectiveness, and efficiency while removing across two widely-used diffusion models and in both conditional and unconditional image generation scenarios. In our experiments, we demonstrate the unlearning of classes, attributes, and even a race from face and object datasets such as UTKFace, CelebA, CelebA-HQ, and CIFAR10.
On the representation and methodology for wide and short range head pose estimation
Authors: Authors: Alejandro Cobo, Roberto Valle, José M. Buenaposada, Luis Baumela
Subjects: Computer Vision and Pattern Recognition (cs.CV)
Abstract
Head pose estimation (HPE) is a problem of interest in computer vision to improve the performance of face processing tasks in semi-frontal or profile settings. Recent applications require the analysis of faces in the full 360{\deg} rotation range. Traditional approaches to solve the semi-frontal and profile cases are not directly amenable for the full rotation case. In this paper we analyze the methodology for short- and wide-range HPE and discuss which representations and metrics are adequate for each case. We show that the popular Euler angles representation is a good choice for short-range HPE, but not at extreme rotations. However, the Euler angles' gimbal lock problem prevents them from being used as a valid metric in any setting. We also revisit the current cross-data set evaluation methodology and note that the lack of alignment between the reference systems of the training and test data sets negatively biases the results of all articles in the literature. We introduce a procedure to quantify this misalignment and a new methodology for cross-data set HPE that establishes new, more accurate, SOTA for the 300W-LP|Biwi benchmark. We also propose a generalization of the geodesic angular distance metric that enables the construction of a loss that controls the contribution of each training sample to the optimization of the model. Finally, we introduce a wide range HPE benchmark based on the CMU Panoptic data set.
Privacy Analysis of Affine Transformations in Cloud-based MPC: Vulnerability to Side-knowledge
Authors: Authors: Teimour Hosseinalizadeh, Nils Schlüter, Moritz Schulze Darup, Nima Monshizadeh
Abstract
Search for the optimizer in computationally demanding model predictive control (MPC) setups can be facilitated by Cloud as a service provider in cyber-physical systems. This advantage introduces the risk that Cloud can obtain unauthorized access to the privacy-sensitive parameters of the system and cost function. To solve this issue, i.e., preventing Cloud from accessing the parameters while benefiting from Cloud computation, random affine transformations provide an exact yet light weight in computation solution. This research deals with analyzing privacy preserving properties of these transformations when they are adopted for MPC problems. We consider two common strategies for outsourcing the optimization required in MPC problems, namely separate and dense forms, and establish that random affine transformations utilized in these forms are vulnerable to side-knowledge from Cloud. Specifically, we prove that the privacy guarantees of these methods and their extensions for separate form are undermined when a mild side-knowledge about the problem in terms of structure of MPC cost function is available. In addition, while we prove that outsourcing the MPC problem in the dense form inherently leads to some degree of privacy for the system and cost function parameters, we also establish that affine transformations applied to this form are nevertheless prone to be undermined by a Cloud with mild side-knowledge. Numerical simulations confirm our results.
On State Estimation in Multi-Sensor Fusion Navigation: Optimization and Filtering
Abstract
The essential of navigation, perception, and decision-making which are basic tasks for intelligent robots, is to estimate necessary system states. Among them, navigation is fundamental for other upper applications, providing precise position and orientation, by integrating measurements from multiple sensors. With observations of each sensor appropriately modelled, multi-sensor fusion tasks for navigation are reduced to the state estimation problem which can be solved by two approaches: optimization and filtering. Recent research has shown that optimization-based frameworks outperform filtering-based ones in terms of accuracy. However, both methods are based on maximum likelihood estimation (MLE) and should be theoretically equivalent with the same linearization points, observation model, measurements, and Gaussian noise assumption. In this paper, we deeply dig into the theories and existing strategies utilized in both optimization-based and filtering-based approaches. It is demonstrated that the two methods are equal theoretically, but this equivalence corrupts due to different strategies applied in real-time operation. By adjusting existing strategies of the filtering-based approaches, the Monte-Carlo simulation and vehicular ablation experiments based on visual odometry (VO) indicate that the strategy adjusted filtering strictly equals to optimization. Therefore, future research on sensor-fusion problems should concentrate on their own algorithms and strategies rather than state estimation approaches.
Secure Dynamic Event-triggered Consensus Under Asynchronous Denial of Service
Abstract
This article proposes a secure implementation for consensus using a dynamic event-triggered (DET) communication scheme in high-order nonlinear multi-agent systems (MAS) under asynchronous (distributed) denial of service (DoS) attacks. By introducing a linear auxiliary trajectory of the system, the DET data transmission scheme among the neighboring agents is employed to reduce the communication for each agent. The asynchronous DoS attacks can block each communication channel among the cooperative agents independently in an unknown pattern. To guarantee state consensus of auxiliary MAS under DoS, a linear matrix inequality (LMI) based optimization approach is proposed which simultaneously designs all the unknown DET communication parameters as well as the state feedback control gain. In addition to asynchronous DoS attacks over the graph topology, the destructive effects of independent DoS attacks over the communication links between actual and auxiliary states are compensated as an additional layer of resiliency for the system. The output of each agent ultimately tracks the auxiliary state of the system and this results in the output consensus.
Optimistic Model Rollouts for Pessimistic Offline Policy Optimization
Authors: Authors: Yuanzhao Zhai, Yiying Li, Zijian Gao, Xudong Gong, Kele Xu, Dawei Feng, Ding Bo, Huaimin Wang
Abstract
Model-based offline reinforcement learning (RL) has made remarkable progress, offering a promising avenue for improving generalization with synthetic model rollouts. Existing works primarily focus on incorporating pessimism for policy optimization, usually via constructing a Pessimistic Markov Decision Process (P-MDP). However, the P-MDP discourages the policies from learning in out-of-distribution (OOD) regions beyond the support of offline datasets, which can under-utilize the generalization ability of dynamics models. In contrast, we propose constructing an Optimistic MDP (O-MDP). We initially observed the potential benefits of optimism brought by encouraging more OOD rollouts. Motivated by this observation, we present ORPO, a simple yet effective model-based offline RL framework. ORPO generates Optimistic model Rollouts for Pessimistic offline policy Optimization. Specifically, we train an optimistic rollout policy in the O-MDP to sample more OOD model rollouts. Then we relabel the sampled state-action pairs with penalized rewards and optimize the output policy in the P-MDP. Theoretically, we demonstrate that the performance of policies trained with ORPO can be lower-bounded in linear MDPs. Experimental results show that our framework significantly outperforms P-MDP baselines by a margin of 30%, achieving state-of-the-art performance on the widely-used benchmark. Moreover, ORPO exhibits notable advantages in problems that require generalization.
A k-swap Local Search for Makespan Scheduling
Authors: Authors: Lars Rohwedder, Ashkan Safari, Tjark Vredeveld
Abstract
Local search is a widely used technique for tackling challenging optimization problems, offering significant advantages in terms of computational efficiency and exhibiting strong empirical behavior across a wide range of problem domains. In this paper, we address a scheduling problem on two identical parallel machines with the objective of \emph{makespan minimization}. For this problem, we consider a local search neighborhood, called \emph{$k$-swap}, which is a more generalized version of the widely-used \emph{swap} and \emph{jump} neighborhoods. The $k$-swap neighborhood is obtained by swapping at most $k$ jobs between two machines in our schedule. First, we propose an algorithm for finding an improving neighbor in the $k$-swap neighborhood which is faster than the naive approach, and prove an almost matching lower bound on any such an algorithm. Then, we analyze the number of local search steps required to converge to a local optimum with respect to the $k$-swap neighborhood. For the case $k = 2$ (similar to the swap neighborhood), we provide a polynomial upper bound on the number of local search steps, and for the case $k = 3$, we provide an exponential lower bound. Finally, we conduct computational experiments on various families of instances, and we discuss extensions to more than two machines in our schedule.
Machine Learning Insides OptVerse AI Solver: Design Principles and Applications
Authors: Authors: Xijun Li, Fangzhou Zhu, Hui-Ling Zhen, Weilin Luo, Meng Lu, Yimin Huang, Zhenan Fan, Zirui Zhou, Yufei Kuang, Zhihai Wang, Zijie Geng, Yang Li, Haoyang Liu, Zhiwu An, Muming Yang, Jianshu Li, Jie Wang, Junchi Yan, Defeng Sun, Tao Zhong, Yong Zhang, Jia Zeng, Mingxuan Yuan, Jianye Hao, Jun Yao, Kun Mao
Abstract
In an era of digital ubiquity, efficient resource management and decision-making are paramount across numerous industries. To this end, we present a comprehensive study on the integration of machine learning (ML) techniques into Huawei Cloud's OptVerse AI Solver, which aims to mitigate the scarcity of real-world mathematical programming instances, and to surpass the capabilities of traditional optimization techniques. We showcase our methods for generating complex SAT and MILP instances utilizing generative models that mirror multifaceted structures of real-world problem. Furthermore, we introduce a training framework leveraging augmentation policies to maintain solvers' utility in dynamic environments. Besides the data generation and augmentation, our proposed approaches also include novel ML-driven policies for personalized solver strategies, with an emphasis on applications like graph convolutional networks for initial basis selection and reinforcement learning for advanced presolving and cut selection. Additionally, we detail the incorporation of state-of-the-art parameter tuning algorithms which markedly elevate solver performance. Compared with traditional solvers such as Gurobi and SCIP, our ML-augmented OptVerse AI Solver demonstrates superior speed and precision across both established benchmarks and real-world scenarios, reinforcing the practical imperative and effectiveness of machine learning techniques in mathematical programming solvers.
HAP: SPMD DNN Training on Heterogeneous GPU Clusters with Automated Program Synthesis
Abstract
Single-Program-Multiple-Data (SPMD) parallelism has recently been adopted to train large deep neural networks (DNNs). Few studies have explored its applicability on heterogeneous clusters, to fully exploit available resources for large model learning. This paper presents \OurSystem, an automated system designed to expedite SPMD DNN training on heterogeneous clusters. \OurSystem jointly optimizes the tensor sharding strategy, sharding ratios across heterogeneous devices and the communication methods for tensor exchanges for optimized distributed training with SPMD parallelism. We novelly formulate model partitioning as a program synthesis problem, in which we generate a distributed program from scratch on a distributed instruction set that semantically resembles the program designed for a single device, and systematically explore the solution space with an A*-based search algorithm. We derive the optimal tensor sharding ratios by formulating it as a linear programming problem. Additionally, \OurSystem explores tensor communication optimization in a heterogeneous cluster and integrates it as part of the program synthesis process, for automatically choosing optimal collective communication primitives and applying sufficient factor broadcasting technique. Extensive experiments on representative workloads demonstrate that \OurSystem achieves up to 2.41x speed-up on heterogeneous clusters.
Block-Diagonal Orthogonal Relation and Matrix Entity for Knowledge Graph Embedding
Authors: Authors: Yihua Zhu, Hidetoshi Shimodaira
Subjects: Computation and Language (cs.CL); Information Retrieval (cs.IR)
Abstract
The primary aim of Knowledge Graph embeddings (KGE) is to learn low-dimensional representations of entities and relations for predicting missing facts. While rotation-based methods like RotatE and QuatE perform well in KGE, they face two challenges: limited model flexibility requiring proportional increases in relation size with entity dimension, and difficulties in generalizing the model for higher-dimensional rotations. To address these issues, we introduce OrthogonalE, a novel KGE model employing matrices for entities and block-diagonal orthogonal matrices with Riemannian optimization for relations. This approach enhances the generality and flexibility of KGE models. The experimental results indicate that our new KGE model, OrthogonalE, is both general and flexible, significantly outperforming state-of-the-art KGE models while substantially reducing the number of relation parameters.
Spatial-Aware Deep Reinforcement Learning for the Traveling Officer Problem
Abstract
The traveling officer problem (TOP) is a challenging stochastic optimization task. In this problem, a parking officer is guided through a city equipped with parking sensors to fine as many parking offenders as possible. A major challenge in TOP is the dynamic nature of parking offenses, which randomly appear and disappear after some time, regardless of whether they have been fined. Thus, solutions need to dynamically adjust to currently fineable parking offenses while also planning ahead to increase the likelihood that the officer arrives during the offense taking place. Though various solutions exist, these methods often struggle to take the implications of actions on the ability to fine future parking violations into account. This paper proposes SATOP, a novel spatial-aware deep reinforcement learning approach for TOP. Our novel state encoder creates a representation of each action, leveraging the spatial relationships between parking spots, the agent, and the action. Furthermore, we propose a novel message-passing module for learning future inter-action correlations in the given environment. Thus, the agent can estimate the potential to fine further parking violations after executing an action. We evaluate our method using an environment based on real-world data from Melbourne. Our results show that SATOP consistently outperforms state-of-the-art TOP agents and is able to fine up to 22% more parking offenses.
End-to-end Learnable Clustering for Intent Learning in Recommendation
Authors: Authors: Yue Liu, Shihao Zhu, Jun Xia, Yingwei Ma, Jian Ma, Wenliang Zhong, Guannan Zhang, Kejun Zhang, Xinwang Liu
Subjects: Information Retrieval (cs.IR); Artificial Intelligence (cs.AI)
Abstract
Mining users' intents plays a crucial role in sequential recommendation. The recent approach, ICLRec, was introduced to extract underlying users' intents using contrastive learning and clustering. While it has shown effectiveness, the existing method suffers from complex and cumbersome alternating optimization, leading to two main issues. Firstly, the separation of representation learning and clustering optimization within a generalized expectation maximization (EM) framework often results in sub-optimal performance. Secondly, performing clustering on the entire dataset hampers scalability for large-scale industry data. To address these challenges, we propose a novel intent learning method called \underline{ELCRec}, which integrates representation learning into an \underline{E}nd-to-end \underline{L}earnable \underline{C}lustering framework for \underline{Rec}ommendation. Specifically, we encode users' behavior sequences and initialize the cluster centers as learnable network parameters. Additionally, we design a clustering loss that guides the networks to differentiate between different cluster centers and pull similar samples towards their respective cluster centers. This allows simultaneous optimization of recommendation and clustering using mini-batch data. Moreover, we leverage the learned cluster centers as self-supervision signals for representation learning, resulting in further enhancement of recommendation performance. Extensive experiments conducted on open benchmarks and industry data validate the superiority, effectiveness, and efficiency of our proposed ELCRec method. Code is available at: https://github.com/yueliu1999/ELCRec.
HybridOctree_Hex: Hybrid Octree-Based Adaptive All-Hexahedral Mesh Generation with Jacobian Control
Authors: Authors: Hua Tong, Eni Halilaj, Yongjie Jessica Zhang
Abstract
We present a new software package "HybridOctree_Hex" for adaptive all-hexahedral mesh generation based on hybrid octree and quality improvement with Jacobian control. The proposed HybridOctree_Hex begins by detecting curvatures and narrow regions of the input boundary to identify key surface features and initialize an octree structure. Subsequently, a strongly balanced octree is constructed using the balancing and pairing rules. Inspired by our earlier preliminary hybrid octree-based work, templates are designed to guarantee an all-hexahedral dual mesh generation directly from the strongly balanced octree. With these pre-defined templates, the sophisticated hybrid octree construction step is skipped to achieve an efficient implementation. After that, elements outside and around the boundary are removed to create a core mesh. The boundary points of the core mesh are connected to their corresponding closest points on the surface to fill the buffer zone and build the final mesh. Coupled with smart Laplacian smoothing, HybridOctree_Hex takes advantage of a delicate optimization-based quality improvement method considering geometric fitting, Jacobian and scaled Jacobian to achieve the minimum scaled Jacobian higher than $0.5$. We empirically verify the robustness and efficiency of our method by running the HybridOctree_Hex software on dozens of complex 3D models without any manual intervention or parameter adjustment. We provide the HybridOctree_Hex source codes, comprehensive results, encompassing mesh input/output files and statistical data presented at https://github.com/CMU-CBML/HybridOctree_Hex.
An Opportunistic Source Synthesis Method for Smart Electromagnetic Environments
Authors: Authors: Pietro Da Rù, Arianna Benoni, Marco Salucci, Paolo Rocca, Andrea Massa
Abstract
In the framework of the "Smart ElectroMagnetic Environment" (SEME), an innovative strategy leveraging Equivalence Source concepts is introduced for enhancing the performance of large-scale outdoor wireless communication systems. The proposed Opportunistic Sources Synthesis (OSS) approach is aimed at unconventionally synthesizing the primary source (i.e., the base transceiver station (BTS) antenna array), so that the complex scattering phenomena induced in the surrounding scatterers are profitably exploited to enhance the received power within user-defined regions of interest (RoIs). To yield a computationally feasible synthesis process, an innovative "Embedded-plus-Environment Patterns" (EPEPs) method is introduced. A set of representative numerical examples, concerned with realistic large-scale outdoor scenarios, is presented to assess the effectiveness and the efficiency of the proposed optimization-driven approach for a realistic SEME implementation.
TRIPS: Trilinear Point Splatting for Real-Time Radiance Field Rendering
Authors: Authors: Linus Franke, Darius Rückert, Laura Fink, Marc Stamminger
Subjects: Computer Vision and Pattern Recognition (cs.CV); Graphics (cs.GR)
Abstract
Point-based radiance field rendering has demonstrated impressive results for novel view synthesis, offering a compelling blend of rendering quality and computational efficiency. However, also latest approaches in this domain are not without their shortcomings. 3D Gaussian Splatting [Kerbl and Kopanas et al. 2023] struggles when tasked with rendering highly detailed scenes, due to blurring and cloudy artifacts. On the other hand, ADOP [R\"uckert et al. 2022] can accommodate crisper images, but the neural reconstruction network decreases performance, it grapples with temporal instability and it is unable to effectively address large gaps in the point cloud. In this paper, we present TRIPS (Trilinear Point Splatting), an approach that combines ideas from both Gaussian Splatting and ADOP. The fundamental concept behind our novel technique involves rasterizing points into a screen-space image pyramid, with the selection of the pyramid layer determined by the projected point size. This approach allows rendering arbitrarily large points using a single trilinear write. A lightweight neural network is then used to reconstruct a hole-free image including detail beyond splat resolution. Importantly, our render pipeline is entirely differentiable, allowing for automatic optimization of both point sizes and positions. Our evaluation demonstrate that TRIPS surpasses existing state-of-the-art methods in terms of rendering quality while maintaining a real-time frame rate of 60 frames per second on readily available hardware. This performance extends to challenging scenarios, such as scenes featuring intricate geometry, expansive landscapes, and auto-exposed footage.
Topology-Driven Parallel Trajectory Optimization in Dynamic Environments
Authors: Authors: Oscar de Groot, Laura Ferranti, Dariu Gavrila, Javier Alonso-Mora
Abstract
Ground robots navigating in complex, dynamic environments must compute collision-free trajectories to avoid obstacles safely and efficiently. Nonconvex optimization is a popular method to compute a trajectory in real-time. However, these methods often converge to locally optimal solutions and frequently switch between different local minima, leading to inefficient and unsafe robot motion. In this work, We propose a novel topology-driven trajectory optimization strategy for dynamic environments that plans multiple distinct evasive trajectories to enhance the robot's behavior and efficiency. A global planner iteratively generates trajectories in distinct homotopy classes. These trajectories are then optimized by local planners working in parallel. While each planner shares the same navigation objectives, they are locally constrained to a specific homotopy class, meaning each local planner attempts a different evasive maneuver. The robot then executes the feasible trajectory with the lowest cost in a receding horizon manner. We demonstrate, on a mobile robot navigating among pedestrians, that our approach leads to faster and safer trajectories than existing planners.
Secrets of RLHF in Large Language Models Part II: Reward Modeling
Authors: Authors: Binghai Wang, Rui Zheng, Lu Chen, Yan Liu, Shihan Dou, Caishuang Huang, Wei Shen, Senjie Jin, Enyu Zhou, Chenyu Shi, Songyang Gao, Nuo Xu, Yuhao Zhou, Xiaoran Fan, Zhiheng Xi, Jun Zhao, Xiao Wang, Tao Ji, Hang Yan, Lixing Shen, Zhan Chen, Tao Gui, Qi Zhang, Xipeng Qiu, Xuanjing Huang, Zuxuan Wu, Yu-Gang Jiang
Abstract
Reinforcement Learning from Human Feedback (RLHF) has become a crucial technology for aligning language models with human values and intentions, enabling models to produce more helpful and harmless responses. Reward models are trained as proxies for human preferences to drive reinforcement learning optimization. While reward models are often considered central to achieving high performance, they face the following challenges in practical applications: (1) Incorrect and ambiguous preference pairs in the dataset may hinder the reward model from accurately capturing human intent. (2) Reward models trained on data from a specific distribution often struggle to generalize to examples outside that distribution and are not suitable for iterative RLHF training. In this report, we attempt to address these two issues. (1) From a data perspective, we propose a method to measure the strength of preferences within the data, based on a voting mechanism of multiple reward models. Experimental results confirm that data with varying preference strengths have different impacts on reward model performance. We introduce a series of novel methods to mitigate the influence of incorrect and ambiguous preferences in the dataset and fully leverage high-quality preference data. (2) From an algorithmic standpoint, we introduce contrastive learning to enhance the ability of reward models to distinguish between chosen and rejected responses, thereby improving model generalization. Furthermore, we employ meta-learning to enable the reward model to maintain the ability to differentiate subtle differences in out-of-distribution samples, and this approach can be utilized for iterative RLHF optimization.
Keyword: adam
There is no result
Keyword: gradient
DISTWAR: Fast Differentiable Rendering on Raster-based Rendering Pipelines
Authors: Authors: Sankeerth Durvasula, Adrian Zhao, Fan Chen, Ruofan Liang, Pawan Kumar Sanjaya, Nandita Vijaykumar
Abstract
Differentiable rendering is a technique used in an important emerging class of visual computing applications that involves representing a 3D scene as a model that is trained from 2D images using gradient descent. Recent works (e.g. 3D Gaussian Splatting) use a rasterization pipeline to enable rendering high quality photo-realistic imagery at high speeds from these learned 3D models. These methods have been demonstrated to be very promising, providing state-of-art quality for many important tasks. However, training a model to represent a scene is still a time-consuming task even when using powerful GPUs. In this work, we observe that the gradient computation phase during training is a significant bottleneck on GPUs due to the large number of atomic operations that need to be processed. These atomic operations overwhelm atomic units in the L2 partitions causing stalls. To address this challenge, we leverage the observations that during the gradient computation: (1) for most warps, all threads atomically update the same memory locations; and (2) warps generate varying amounts of atomic traffic (since some threads may be inactive). We propose DISTWAR, a software-approach to accelerate atomic operations based on two key ideas: First, we enable warp-level reduction of threads at the SM sub-cores using registers to leverage the locality in intra-warp atomic updates. Second, we distribute the atomic computation between the warp-level reduction at the SM and the L2 atomic units to increase the throughput of atomic computation. Warps with many threads performing atomic updates to the same memory locations are scheduled at the SM, and the rest using L2 atomic units. We implement DISTWAR using existing warp-level primitives. We evaluate DISTWAR on widely used raster-based differentiable rendering workloads. We demonstrate significant speedups of 2.44x on average (up to 5.7x).
Fully Spiking Actor Network with Intra-layer Connections for Reinforcement Learning
Abstract
With the help of special neuromorphic hardware, spiking neural networks (SNNs) are expected to realize artificial intelligence (AI) with less energy consumption. It provides a promising energy-efficient way for realistic control tasks by combining SNNs with deep reinforcement learning (DRL). In this paper, we focus on the task where the agent needs to learn multi-dimensional deterministic policies to control, which is very common in real scenarios. Recently, the surrogate gradient method has been utilized for training multi-layer SNNs, which allows SNNs to achieve comparable performance with the corresponding deep networks in this task. Most existing spike-based RL methods take the firing rate as the output of SNNs, and convert it to represent continuous action space (i.e., the deterministic policy) through a fully-connected (FC) layer. However, the decimal characteristic of the firing rate brings the floating-point matrix operations to the FC layer, making the whole SNN unable to deploy on the neuromorphic hardware directly. To develop a fully spiking actor network without any floating-point matrix operations, we draw inspiration from the non-spiking interneurons found in insects and employ the membrane voltage of the non-spiking neurons to represent the action. Before the non-spiking neurons, multiple population neurons are introduced to decode different dimensions of actions. Since each population is used to decode a dimension of action, we argue that the neurons in each population should be connected in time domain and space domain. Hence, the intra-layer connections are used in output populations to enhance the representation capacity. Finally, we propose a fully spiking actor network with intra-layer connections (ILC-SAN).
D3GU: Multi-Target Active Domain Adaptation via Enhancing Domain Alignment
Authors: Authors: Lin Zhang, Linghan Xu, Saman Motamed, Shayok Chakraborty, Fernando De la Torre
Subjects: Computer Vision and Pattern Recognition (cs.CV)
Abstract
Unsupervised domain adaptation (UDA) for image classification has made remarkable progress in transferring classification knowledge from a labeled source domain to an unlabeled target domain, thanks to effective domain alignment techniques. Recently, in order to further improve performance on a target domain, many Single-Target Active Domain Adaptation (ST-ADA) methods have been proposed to identify and annotate the salient and exemplar target samples. However, it requires one model to be trained and deployed for each target domain and the domain label associated with each test sample. This largely restricts its application in the ubiquitous scenarios with multiple target domains. Therefore, we propose a Multi-Target Active Domain Adaptation (MT-ADA) framework for image classification, named D3GU, to simultaneously align different domains and actively select samples from them for annotation. This is the first research effort in this field to our best knowledge. D3GU applies Decomposed Domain Discrimination (D3) during training to achieve both source-target and target-target domain alignments. Then during active sampling, a Gradient Utility (GU) score is designed to weight every unlabeled target image by its contribution towards classification and domain alignment tasks, and is further combined with KMeans clustering to form GU-KMeans for diverse image sampling. Extensive experiments on three benchmark datasets, Office31, OfficeHome, and DomainNet, have been conducted to validate consistently superior performance of D3GU for MT-ADA.
Towards Safe Load Balancing based on Control Barrier Functions and Deep Reinforcement Learning
Authors: Authors: Lam Dinh, Pham Tran Anh Quang, Jérémie Leguay
Subjects: Networking and Internet Architecture (cs.NI); Machine Learning (cs.LG)
Abstract
Deep Reinforcement Learning (DRL) algorithms have recently made significant strides in improving network performance. Nonetheless, their practical use is still limited in the absence of safe exploration and safe decision-making. In the context of commercial solutions, reliable and safe-to-operate systems are of paramount importance. Taking this problem into account, we propose a safe learning-based load balancing algorithm for Software Defined-Wide Area Network (SD-WAN), which is empowered by Deep Reinforcement Learning (DRL) combined with a Control Barrier Function (CBF). It safely projects unsafe actions into feasible ones during both training and testing, and it guides learning towards safe policies. We successfully implemented the solution on GPU to accelerate training by approximately 110x times and achieve model updates for on-policy methods within a few seconds, making the solution practical. We show that our approach delivers near-optimal Quality-of-Service (QoS performance in terms of end-to-end delay while respecting safety requirements related to link capacity constraints. We also demonstrated that on-policy learning based on Proximal Policy Optimization (PPO) performs better than off-policy learning with Deep Deterministic Policy Gradient (DDPG) when both are combined with a CBF for safe load balancing.
Optimization by linear kinetic equations and mean-field Langevin dynamics
Abstract
Probably one of the most striking examples of the close connections between global optimization processes and statistical physics is the simulated annealing method, inspired by the famous Monte Carlo algorithm devised by Metropolis et al. in the middle of the last century. In this paper we show how the tools of linear kinetic theory allow to describe this gradient-free algorithm from the perspective of statistical physics and how convergence to the global minimum can be related to classical entropy inequalities. This analysis highlight the strong link between linear Boltzmann equations and stochastic optimization methods governed by Markov processes. Thanks to this formalism we can establish the connections between the simulated annealing process and the corresponding mean-field Langevin dynamics characterized by a stochastic gradient descent approach. Generalizations to other selection strategies in simulated annealing that avoid the acceptance-rejection dynamic are also provided.
GE-AdvGAN: Improving the transferability of adversarial samples by gradient editing-based adversarial generative model
Abstract
Adversarial generative models, such as Generative Adversarial Networks (GANs), are widely applied for generating various types of data, i.e., images, text, and audio. Accordingly, its promising performance has led to the GAN-based adversarial attack methods in the white-box and black-box attack scenarios. The importance of transferable black-box attacks lies in their ability to be effective across different models and settings, more closely aligning with real-world applications. However, it remains challenging to retain the performance in terms of transferable adversarial examples for such methods. Meanwhile, we observe that some enhanced gradient-based transferable adversarial attack algorithms require prolonged time for adversarial sample generation. Thus, in this work, we propose a novel algorithm named GE-AdvGAN to enhance the transferability of adversarial samples whilst improving the algorithm's efficiency. The main approach is via optimising the training process of the generator parameters. With the functional and characteristic similarity analysis, we introduce a novel gradient editing (GE) mechanism and verify its feasibility in generating transferable samples on various models. Moreover, by exploring the frequency domain information to determine the gradient editing direction, GE-AdvGAN can generate highly transferable adversarial samples while minimizing the execution time in comparison to the state-of-the-art transferable adversarial attack algorithms. The performance of GE-AdvGAN is comprehensively evaluated by large-scale experiments on different datasets, which results demonstrate the superiority of our algorithm. The code for our algorithm is available at: https://github.com/LMBTough/GE-advGAN
Keyword: super-resolution
Transforming Image Super-Resolution: A ConvFormer-based Efficient Approach
Authors: Authors: Gang Wu, Junjun Jiang, Junpeng Jiang, Xianming Liu
Subjects: Computer Vision and Pattern Recognition (cs.CV); Image and Video Processing (eess.IV)
Abstract
Recent progress in single-image super-resolution (SISR) has achieved remarkable performance, yet the computational costs of these methods remain a challenge for deployment on resource-constrained devices. Especially for transformer-based methods, the self-attention mechanism in such models brings great breakthroughs while incurring substantial computational costs. To tackle this issue, we introduce the Convolutional Transformer layer (ConvFormer) and the ConvFormer-based Super-Resolution network (CFSR), which offer an effective and efficient solution for lightweight image super-resolution tasks. In detail, CFSR leverages the large kernel convolution as the feature mixer to replace the self-attention module, efficiently modeling long-range dependencies and extensive receptive fields with a slight computational cost. Furthermore, we propose an edge-preserving feed-forward network, simplified as EFN, to obtain local feature aggregation and simultaneously preserve more high-frequency information. Extensive experiments demonstrate that CFSR can achieve an advanced trade-off between computational cost and performance when compared to existing lightweight SR methods. Compared to state-of-the-art methods, e.g. ShuffleMixer, the proposed CFSR achieves 0.39 dB gains on Urban100 dataset for x2 SR task while containing 26% and 31% fewer parameters and FLOPs, respectively. Code and pre-trained models are available at https://github.com/Aitical/CFSR.
Keyword: sgd
Correlated Quantization for Faster Nonconvex Distributed Optimization
Keyword: optimization
Dataset Optimization for Chronic Disease Prediction with Bio-Inspired Feature Selection
Functional Graphical Models: Structure Enables Offline Data-Driven Optimization
Standardizing Your Training Process for Human Activity Recognition Models: A Comprehensive Review in the Tunable Factors
Optimized Ensemble Model Towards Secured Industrial IoT Devices
FPRF: Feed-Forward Photorealistic Style Transfer of Large-Scale 3D Neural Radiance Fields
Correlated Quantization for Faster Nonconvex Distributed Optimization
Towards Safe Load Balancing based on Control Barrier Functions and Deep Reinforcement Learning
Multi-objective Feature Selection in Remote Health Monitoring Applications
Optimization by linear kinetic equations and mean-field Langevin dynamics
Characterising the take-off dynamics and energy efficiency in spring-driven jumping robots
Graph Q-Learning for Combinatorial Optimization
MatSAM: Efficient Materials Microstructure Extraction via Visual Large Model
Parrot: Pareto-optimal Multi-Reward Reinforcement Learning Framework for Text-to-Image Generation
Regularization of the discrete source problem in the nonlinear diffusive-logistic equation
CAT-LLM: Prompting Large Language Models with Text Style Definition for Chinese Article-style Transfer
Kernelized Normalizing Constant Estimation: Bridging Bayesian Quadrature and Bayesian Optimization
STAR-RIS Enhanced UAV-Enabled MEC Networks with Bi-Directional Task Offloading
An experimental evaluation of Deep Reinforcement Learning algorithms for HVAC control
Cross-Inlining Binary Function Similarity Detection
GO-NeRF: Generating Virtual Objects in Neural Radiance Fields
EraseDiff: Erasing Data Influence in Diffusion Models
On the representation and methodology for wide and short range head pose estimation
Privacy Analysis of Affine Transformations in Cloud-based MPC: Vulnerability to Side-knowledge
On State Estimation in Multi-Sensor Fusion Navigation: Optimization and Filtering
Secure Dynamic Event-triggered Consensus Under Asynchronous Denial of Service
Optimistic Model Rollouts for Pessimistic Offline Policy Optimization
A k-swap Local Search for Makespan Scheduling
Machine Learning Insides OptVerse AI Solver: Design Principles and Applications
HAP: SPMD DNN Training on Heterogeneous GPU Clusters with Automated Program Synthesis
Block-Diagonal Orthogonal Relation and Matrix Entity for Knowledge Graph Embedding
Spatial-Aware Deep Reinforcement Learning for the Traveling Officer Problem
End-to-end Learnable Clustering for Intent Learning in Recommendation
HybridOctree_Hex: Hybrid Octree-Based Adaptive All-Hexahedral Mesh Generation with Jacobian Control
An Opportunistic Source Synthesis Method for Smart Electromagnetic Environments
TRIPS: Trilinear Point Splatting for Real-Time Radiance Field Rendering
Topology-Driven Parallel Trajectory Optimization in Dynamic Environments
Secrets of RLHF in Large Language Models Part II: Reward Modeling
Keyword: adam
There is no result
Keyword: gradient
DISTWAR: Fast Differentiable Rendering on Raster-based Rendering Pipelines
Fully Spiking Actor Network with Intra-layer Connections for Reinforcement Learning
D3GU: Multi-Target Active Domain Adaptation via Enhancing Domain Alignment
Towards Safe Load Balancing based on Control Barrier Functions and Deep Reinforcement Learning
Optimization by linear kinetic equations and mean-field Langevin dynamics
GE-AdvGAN: Improving the transferability of adversarial samples by gradient editing-based adversarial generative model
Keyword: super-resolution
Transforming Image Super-Resolution: A ConvFormer-based Efficient Approach