Abstract
Minimally invasive surgeries and related applications demand surgical tool classification and segmentation at the instance level. Surgical tools are similar in appearance and are long, thin, and handled at an angle. The fine-tuning of state-of-the-art (SOTA) instance segmentation models trained on natural images for instrument segmentation has difficulty discriminating instrument classes. Our research demonstrates that while the bounding box and segmentation mask are often accurate, the classification head mis-classifies the class label of the surgical instrument. We present a new neural network framework that adds a classification module as a new stage to existing instance segmentation models. This module specializes in improving the classification of instrument masks generated by the existing model. The module comprises multi-scale mask attention, which attends to the instrument region and masks the distracting background features. We propose training our classifier module using metric learning with arc loss to handle low inter-class variance of surgical instruments. We conduct exhaustive experiments on the benchmark datasets EndoVis2017 and EndoVis2018. We demonstrate that our method outperforms all (more than 18) SOTA methods compared with, and improves the SOTA performance by at least 12 points (20%) on the EndoVis2017 benchmark challenge and generalizes effectively across the datasets.
Advancing Deep Metric Learning Through Multiple Batch Norms And Multi-Targeted Adversarial Examples
Abstract
Deep Metric Learning (DML) is a prominent field in machine learning with extensive practical applications that concentrate on learning visual similarities. It is known that inputs such as Adversarial Examples (AXs), which follow a distribution different from that of clean data, result in false predictions from DML systems. This paper proposes MDProp, a framework to simultaneously improve the performance of DML models on clean data and inputs following multiple distributions. MDProp utilizes multi-distribution data through an AX generation process while leveraging disentangled learning through multiple batch normalization layers during the training of a DML model. MDProp is the first to generate feature space multi-targeted AXs to perform targeted regularization on the training model's denser embedding space regions, resulting in improved embedding space densities contributing to the improved generalization in the trained models. From a comprehensive experimental analysis, we show that MDProp results in up to 2.95% increased clean data Recall@1 scores and up to 2.12 times increased robustness against different input distributions compared to the conventional methods.
Intra-class Adaptive Augmentation with Neighbor Correction for Deep Metric Learning
Abstract
Deep metric learning aims to learn an embedding space, where semantically similar samples are close together and dissimilar ones are repelled against. To explore more hard and informative training signals for augmentation and generalization, recent methods focus on generating synthetic samples to boost metric learning losses. However, these methods just use the deterministic and class-independent generations (e.g., simple linear interpolation), which only can cover the limited part of distribution spaces around original samples. They have overlooked the wide characteristic changes of different classes and can not model abundant intra-class variations for generations. Therefore, generated samples not only lack rich semantics within the certain class, but also might be noisy signals to disturb training. In this paper, we propose a novel intra-class adaptive augmentation (IAA) framework for deep metric learning. We reasonably estimate intra-class variations for every class and generate adaptive synthetic samples to support hard samples mining and boost metric learning losses. Further, for most datasets that have a few samples within the class, we propose the neighbor correction to revise the inaccurate estimations, according to our correlation discovery where similar classes generally have similar variation distributions. Extensive experiments on five benchmarks show our method significantly improves and outperforms the state-of-the-art methods on retrieval performances by 3%-6%. Our code is available at https://github.com/darkpromise98/IAA
Learning and Understanding a Disentangled Feature Representation for Hidden Parameters in Reinforcement Learning
Abstract
Hidden parameters are latent variables in reinforcement learning (RL) environments that are constant over the course of a trajectory. Understanding what, if any, hidden parameters affect a particular environment can aid both the development and appropriate usage of RL systems. We present an unsupervised method to map RL trajectories into a feature space where distance represents the relative difference in system behavior due to hidden parameters. Our approach disentangles the effects of hidden parameters by leveraging a recurrent neural network (RNN) world model as used in model-based RL. First, we alter the standard world model training algorithm to isolate the hidden parameter information in the world model memory. Then, we use a metric learning approach to map the RNN memory into a space with a distance metric approximating a bisimulation metric with respect to the hidden parameters. The resulting disentangled feature space can be used to meaningfully relate trajectories to each other and analyze the hidden parameter. We demonstrate our approach on four hidden parameters across three RL environments. Finally we present two methods to help identify and understand the effects of hidden parameters on systems.
Keyword: image retrieval
SLAN: Self-Locator Aided Network for Cross-Modal Understanding
Abstract
Learning fine-grained interplay between vision and language allows to a more accurate understanding for VisionLanguage tasks. However, it remains challenging to extract key image regions according to the texts for semantic alignments. Most existing works are either limited by textagnostic and redundant regions obtained with the frozen detectors, or failing to scale further due to its heavy reliance on scarce grounding (gold) data to pre-train detectors. To solve these problems, we propose Self-Locator Aided Network (SLAN) for cross-modal understanding tasks without any extra gold data. SLAN consists of a region filter and a region adaptor to localize regions of interest conditioned on different texts. By aggregating cross-modal information, the region filter selects key regions and the region adaptor updates their coordinates with text guidance. With detailed region-word alignments, SLAN can be easily generalized to many downstream tasks. It achieves fairly competitive results on five cross-modal understanding tasks (e.g., 85.7% and 69.2% on COCO image-to-text and text-to-image retrieval, surpassing previous SOTA methods). SLAN also demonstrates strong zero-shot and fine-tuned transferability to two localization tasks.
Keyword: self-supervised
Handling Image and Label Resolution Mismatch in Remote Sensing
Authors: Scott Workman, Armin Hadzic, M. Usman Rafique
Subjects: Computer Vision and Pattern Recognition (cs.CV)
Abstract
Though semantic segmentation has been heavily explored in vision literature, unique challenges remain in the remote sensing domain. One such challenge is how to handle resolution mismatch between overhead imagery and ground-truth label sources, due to differences in ground sample distance. To illustrate this problem, we introduce a new dataset and use it to showcase weaknesses inherent in existing strategies that naively upsample the target label to match the image resolution. Instead, we present a method that is supervised using low-resolution labels (without upsampling), but takes advantage of an exemplar set of high-resolution labels to guide the learning process. Our method incorporates region aggregation, adversarial learning, and self-supervised pretraining to generate fine-grained predictions, without requiring high-resolution annotations. Extensive experiments demonstrate the real-world applicability of our approach.
Survey on Self-Supervised Multimodal Representation Learning and Foundation Models
Authors: Sushil Thapa
Subjects: Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Computer Vision and Pattern Recognition (cs.CV); Computer Science and Game Theory (cs.GT)
Abstract
Deep learning has been the subject of growing interest in recent years. Specifically, a specific type called Multimodal learning has shown great promise for solving a wide range of problems in domains such as language, vision, audio, etc. One promising research direction to improve this further has been learning rich and robust low-dimensional data representation of the high-dimensional world with the help of large-scale datasets present on the internet. Because of its potential to avoid the cost of annotating large-scale datasets, self-supervised learning has been the de facto standard for this task in recent years. This paper summarizes some of the landmark research papers that are directly or indirectly responsible to build the foundation of multimodal self-supervised learning of representation today. The paper goes over the development of representation learning over the last few years for each modality and how they were combined to get a multimodal agent later.
MoDA: Map style transfer for self-supervised Domain Adaptation of embodied agents
Authors: Eun Sun Lee, Junho Kim, SangWon Park, Young Min Kim
Subjects: Robotics (cs.RO); Computer Vision and Pattern Recognition (cs.CV)
Abstract
We propose a domain adaptation method, MoDA, which adapts a pretrained embodied agent to a new, noisy environment without ground-truth supervision. Map-based memory provides important contextual information for visual navigation, and exhibits unique spatial structure mainly composed of flat walls and rectangular obstacles. Our adaptation approach encourages the inherent regularities on the estimated maps to guide the agent to overcome the prevalent domain discrepancy in a novel environment. Specifically, we propose an efficient learning curriculum to handle the visual and dynamics corruptions in an online manner, self-supervised with pseudo clean maps generated by style transfer networks. Because the map-based representation provides spatial knowledge for the agent's policy, our formulation can deploy the pretrained policy networks from simulators in a new setting. We evaluate MoDA in various practical scenarios and show that our proposed method quickly enhances the agent's performance in downstream tasks including localization, mapping, exploration, and point-goal navigation.
Textual Enhanced Contrastive Learning for Solving Math Word Problems
Abstract
Solving math word problems is the task that analyses the relation of quantities and requires an accurate understanding of contextual natural language information. Recent studies show that current models rely on shallow heuristics to predict solutions and could be easily misled by small textual perturbations. To address this problem, we propose a Textual Enhanced Contrastive Learning framework, which enforces the models to distinguish semantically similar examples while holding different mathematical logic. We adopt a self-supervised manner strategy to enrich examples with subtle textual variance by textual reordering or problem re-construction. We then retrieve the hardest to differentiate samples from both equation and textual perspectives and guide the model to learn their representations. Experimental results show that our method achieves state-of-the-art on both widely used benchmark datasets and also exquisitely designed challenge datasets in English and Chinese. \footnote{Our code and data is available at \url{https://github.com/yiyunya/Textual_CL_MWP}
Model Extraction Attack against Self-supervised Speech Models
Authors: Tsu-Yuan Hsu, Chen-An Li, Tung-Yu Wu, Hung-yi Lee
Subjects: Sound (cs.SD); Computation and Language (cs.CL); Cryptography and Security (cs.CR); Machine Learning (cs.LG); Audio and Speech Processing (eess.AS)
Abstract
Self-supervised learning (SSL) speech models generate meaningful representations of given clips and achieve incredible performance across various downstream tasks. Model extraction attack (MEA) often refers to an adversary stealing the functionality of the victim model with only query access. In this work, we study the MEA problem against SSL speech model with a small number of queries. We propose a two-stage framework to extract the model. In the first stage, SSL is conducted on the large-scale unlabeled corpus to pre-train a small speech model. Secondly, we actively sample a small portion of clips from the unlabeled corpus and query the target model with these clips to acquire their representations as labels for the small model's second-stage training. Experiment results show that our sampling methods can effectively extract the target model without knowing any information about its model architecture.
G-CMP: Graph-enhanced Contextual Matrix Profile for unsupervised anomaly detection in sensor-based remote health monitoring
Authors: Nivedita Bijlani, Oscar Mendez Maldonado, Samaneh Kouchaki
Abstract
Sensor-based remote health monitoring is used in industrial, urban and healthcare settings to monitor ongoing operation of equipment and human health. An important aim is to intervene early if anomalous events or adverse health is detected. In the wild, these anomaly detection approaches are challenged by noise, label scarcity, high dimensionality, explainability and wide variability in operating environments. The Contextual Matrix Profile (CMP) is a configurable 2-dimensional version of the Matrix Profile (MP) that uses the distance matrix of all subsequences of a time series to discover patterns and anomalies. The CMP is shown to enhance the effectiveness of the MP and other SOTA methods at detecting, visualising and interpreting true anomalies in noisy real world data from different domains. It excels at zooming out and identifying temporal patterns at configurable time scales. However, the CMP does not address cross-sensor information, and cannot scale to high dimensional data. We propose a novel, self-supervised graph-based approach for temporal anomaly detection that works on context graphs generated from the CMP distance matrix. The learned graph embeddings encode the anomalous nature of a time context. In addition, we evaluate other graph outlier algorithms for the same task. Given our pipeline is modular, graph construction, generation of graph embeddings, and pattern recognition logic can all be chosen based on the specific pattern detection application. We verified the effectiveness of graph-based anomaly detection and compared it with the CMP and 3 state-of-the art methods on two real-world healthcare datasets with different anomalies. Our proposed method demonstrated better recall, alert rate and generalisability.
SgVA-CLIP: Semantic-guided Visual Adapting of Vision-Language Models for Few-shot Image Classification
Authors: Fang Peng, Xiaoshan Yang, Changsheng Xu
Subjects: Computer Vision and Pattern Recognition (cs.CV); Multimedia (cs.MM)
Abstract
Although significant progress has been made in few-shot learning, most of existing few-shot learning methods require supervised pre-training on a large amount of samples of base classes, which limits their generalization ability in real world application. Recently, large-scale self-supervised vision-language models (e.g., CLIP) have provided a new paradigm for transferable visual representation learning. However, the pre-trained VLPs may neglect detailed visual information that is difficult to describe by language sentences, but important for learning an effective classifier in few-shot classification. To address the above problem, we propose a new framework, named Semantic-guided Visual Adapting (SgVA), which can effectively extend vision-language pre-trained models to produce discriminative task-specific visual features by comprehensively using a vision-specific contrastive loss, a cross-modal contrastive loss, and an implicit knowledge distillation. The implicit knowledge distillation is designed to transfer the fine-grained cross-modal knowledge to guide the updating of the vision adapter. State-of-the-art results on 13 datasets demonstrate that the adapted visual features can well complement the cross-modal features to improve few-shot image classification.
Exploring adaptation of VideoMAE for Audio-Visual Diarization & Social @ Ego4d Looking at me Challenge
Authors: Yinan He, Guo Chen
Subjects: Computer Vision and Pattern Recognition (cs.CV)
Abstract
In this report, we present the transferring pretrained video mask autoencoders(VideoMAE) to egocentric tasks for Ego4d Looking at me Challenge. VideoMAE is the data-efficient pretraining model for self-supervised video pre-training and can easily transfer to downstream tasks. We show that the representation transferred from VideoMAE has good Spatio-temporal modeling and the ability to capture small actions. We only need to use egocentric data to train 10 epochs based on VideoMAE which pretrained by the ordinary videos acquired from a third person's view, and we can get better results than the baseline on Ego4d Looking at me Challenge.
Abstract
With infinitely many high-quality data points, infinite computational power, an infinitely large foundation model with a perfect training algorithm and guaranteed zero generalization error on the pretext task, can the model be used for everything? This question cannot be answered by the existing theory of representation, optimization or generalization, because the issues they mainly investigate are assumed to be nonexistent here. In this paper, we show that category theory provides powerful machinery to answer this question. We have proved three results. The first one limits the power of prompt-based learning, saying that the model can solve a downstream task with prompts if and only if the task is representable. The second one says fine tuning does not have this limit, as a foundation model with the minimum power (up to symmetry) can theoretically solve downstream tasks with fine tuning and enough resources. Our final result can be seen as a new type of generalization theorem, showing that the foundation model can generate unseen objects from the target category (e.g., images) using the structural information from the source category (e.g., texts). Along the way, we provide a categorical framework for supervised and self-supervised learning, which might be of independent interest.
BARTSmiles: Generative Masked Language Models for Molecular Representations
Abstract
We discover a robust self-supervised strategy tailored towards molecular representations for generative masked language models through a series of tailored, in-depth ablations. Using this pre-training strategy, we train BARTSmiles, a BART-like model with an order of magnitude more compute than previous self-supervised molecular representations. In-depth evaluations show that BARTSmiles consistently outperforms other self-supervised representations across classification, regression, and generation tasks setting a new state-of-the-art on 11 tasks. We then quantitatively show that when applied to the molecular domain, the BART objective learns representations that implicitly encode our downstream tasks of interest. For example, by selecting seven neurons from a frozen BARTSmiles, we can obtain a model having performance within two percentage points of the full fine-tuned model on task Clintox. Lastly, we show that standard attribution interpretability methods, when applied to BARTSmiles, highlight certain substructures that chemists use to explain specific properties of molecules. The code and the pretrained model are publicly available.
Keyword: vision transformer
LUMix: Improving Mixup by Better Modelling Label Uncertainty
Authors: Shuyang Sun, Jie-Neng Chen, Ruifei He, Alan Yuille, Philip Torr, Song Bai
Subjects: Computer Vision and Pattern Recognition (cs.CV); Machine Learning (cs.LG)
Abstract
Modern deep networks can be better generalized when trained with noisy samples and regularization techniques. Mixup and CutMix have been proven to be effective for data augmentation to help avoid overfitting. Previous Mixup-based methods linearly combine images and labels to generate additional training data. However, this is problematic if the object does not occupy the whole image as we demonstrate in Figure 1. Correctly assigning the label weights is hard even for human beings and there is no clear criterion to measure it. To tackle this problem, in this paper, we propose LUMix, which models such uncertainty by adding label perturbation during training. LUMix is simple as it can be implemented in just a few lines of code and can be universally applied to any deep networks \eg CNNs and Vision Transformers, with minimal computational cost. Extensive experiments show that our LUMix can consistently boost the performance for networks with a wide range of diversity and capacity on ImageNet, \eg $+0.7\%$ for a small model DeiT-S and $+0.6\%$ for a large variant XCiT-L. We also demonstrate that LUMix can lead to better robustness when evaluated on ImageNet-O and ImageNet-A. The source code can be found \href{https://github.com/kevin-ssy/LUMix}{here}
NoisyQuant: Noisy Bias-Enhanced Post-Training Activation Quantization for Vision Transformers
Authors: Yijiang Liu, Huanrui Yang, Zhen Dong, Kurt Keutzer, Li Du, Shanghang Zhang
Subjects: Computer Vision and Pattern Recognition (cs.CV)
Abstract
The complicated architecture and high training cost of vision transformers urge the exploration of post-training quantization. However, the heavy-tailed distribution of vision transformer activations hinders the effectiveness of previous post-training quantization methods, even with advanced quantizer designs. Instead of tuning the quantizer to better fit the complicated activation distribution, this paper proposes NoisyQuant, a quantizer-agnostic enhancement for the post-training activation quantization performance of vision transformers. We make a surprising theoretical discovery that for a given quantizer, adding a fixed Uniform noisy bias to the values being quantized can significantly reduce the quantization error under provable conditions. Building on the theoretical insight, NoisyQuant achieves the first success on actively altering the heavy-tailed activation distribution with additive noisy bias to fit a given quantizer. Extensive experiments show NoisyQuant largely improves the post-training quantization performance of vision transformer with minimal computation overhead. For instance, on linear uniform 6-bit activation quantization, NoisyQuant improves SOTA top-1 accuracy on ImageNet by up to 1.7%, 1.1% and 0.5% for ViT, DeiT, and Swin Transformer respectively, achieving on-par or even higher performance than previous nonlinear, mixed-precision quantization.
Metal-conscious Embedding for CBCT Projection Inpainting
Authors: Fuxin Fan, Yangkong Wang, Ludwig Ritschl, Ramyar Biniazan, Marcel Beister, Björn Kreher, Yixing Huang, Steffen Kappler, Andreas Maier
Subjects: Computer Vision and Pattern Recognition (cs.CV); Image and Video Processing (eess.IV)
Abstract
The existence of metallic implants in projection images for cone-beam computed tomography (CBCT) introduces undesired artifacts which degrade the quality of reconstructed images. In order to reduce metal artifacts, projection inpainting is an essential step in many metal artifact reduction algorithms. In this work, a hybrid network combining the shift window (Swin) vision transformer (ViT) and a convolutional neural network is proposed as a baseline network for the inpainting task. To incorporate metal information for the Swin ViT-based encoder, metal-conscious self-embedding and neighborhood-embedding methods are investigated. Both methods have improved the performance of the baseline network. Furthermore, by choosing appropriate window size, the model with neighborhood-embedding could achieve the lowest mean absolute error of 0.079 in metal regions and the highest peak signal-to-noise ratio of 42.346 in CBCT projections. At the end, the efficiency of metal-conscious embedding on both simulated and real cadaver CBCT data has been demonstrated, where the inpainting capability of the baseline network has been enhanced.
Lightweight Structure-Aware Attention for Visual Understanding
Authors: Heeseung Kwon, Francisco M. Castro, Manuel J. Marin-Jimenez, Nicolas Guil, Karteek Alahari
Subjects: Computer Vision and Pattern Recognition (cs.CV)
Abstract
Vision Transformers (ViTs) have become a dominant paradigm for visual representation learning with self-attention operators. Although these operators provide flexibility to the model with their adjustable attention kernels, they suffer from inherent limitations: (1) the attention kernel is not discriminative enough, resulting in high redundancy of the ViT layers, and (2) the complexity in computation and memory is quadratic in the sequence length. In this paper, we propose a novel attention operator, called lightweight structure-aware attention (LiSA), which has a better representation power with log-linear complexity. Our operator learns structural patterns by using a set of relative position embeddings (RPEs). To achieve log-linear complexity, the RPEs are approximated with fast Fourier transforms. Our experiments and ablation studies demonstrate that ViTs based on the proposed operator outperform self-attention and other existing operators, achieving state-of-the-art results on ImageNet, and competitive results on other visual understanding benchmarks such as COCO and Something-Something-V2. The source code of our approach will be released online.
RGB no more: Minimally-decoded JPEG Vision Transformers
Authors: Jeongsoo Park, Justin Johnson
Subjects: Computer Vision and Pattern Recognition (cs.CV); Image and Video Processing (eess.IV)
Abstract
Most neural networks for computer vision are designed to infer using RGB images. However, these RGB images are commonly encoded in JPEG before saving to disk; decoding them imposes an unavoidable overhead for RGB networks. Instead, our work focuses on training Vision Transformers (ViT) directly from the encoded features of JPEG. This way, we can avoid most of the decoding overhead, accelerating data load. Existing works have studied this aspect but they focus on CNNs. Due to how these encoded features are structured, CNNs require heavy modification to their architecture to accept such data. Here, we show that this is not the case for ViTs. In addition, we tackle data augmentation directly on these encoded features, which to our knowledge, has not been explored in-depth for training in this setting. With these two improvements -- ViT and data augmentation -- we show that our ViT-Ti model achieves up to 39.2% faster training and 17.9% faster inference with no accuracy loss compared to the RGB counterpart.
Finding Differences Between Transformers and ConvNets Using Counterfactual Simulation Testing
Authors: Nataniel Ruiz, Sarah Adel Bargal, Cihang Xie, Kate Saenko, Stan Sclaroff
Abstract
Modern deep neural networks tend to be evaluated on static test sets. One shortcoming of this is the fact that these deep neural networks cannot be easily evaluated for robustness issues with respect to specific scene variations. For example, it is hard to study the robustness of these networks to variations of object scale, object pose, scene lighting and 3D occlusions. The main reason is that collecting real datasets with fine-grained naturalistic variations of sufficient scale can be extremely time-consuming and expensive. In this work, we present Counterfactual Simulation Testing, a counterfactual framework that allows us to study the robustness of neural networks with respect to some of these naturalistic variations by building realistic synthetic scenes that allow us to ask counterfactual questions to the models, ultimately providing answers to questions such as "Would your classification still be correct if the object were viewed from the top?" or "Would your classification still be correct if the object were partially occluded by another object?". Our method allows for a fair comparison of the robustness of recently released, state-of-the-art Convolutional Neural Networks and Vision Transformers, with respect to these naturalistic variations. We find evidence that ConvNext is more robust to pose and scale variations than Swin, that ConvNext generalizes better to our simulated domain and that Swin handles partial occlusion better than ConvNext. We also find that robustness for all networks improves with network scale and with data scale and variety. We release the Naturalistic Variation Object Dataset (NVD), a large simulated dataset of 272k images of everyday objects with naturalistic variations such as object pose, scale, viewpoint, lighting and occlusions. Project page: https://counterfactualsimulation.github.io
Keyword: multimodal
Survey on Self-Supervised Multimodal Representation Learning and Foundation Models
Authors: Sushil Thapa
Subjects: Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Computer Vision and Pattern Recognition (cs.CV); Computer Science and Game Theory (cs.GT)
Abstract
Deep learning has been the subject of growing interest in recent years. Specifically, a specific type called Multimodal learning has shown great promise for solving a wide range of problems in domains such as language, vision, audio, etc. One promising research direction to improve this further has been learning rich and robust low-dimensional data representation of the high-dimensional world with the help of large-scale datasets present on the internet. Because of its potential to avoid the cost of annotating large-scale datasets, self-supervised learning has been the de facto standard for this task in recent years. This paper summarizes some of the landmark research papers that are directly or indirectly responsible to build the foundation of multimodal self-supervised learning of representation today. The paper goes over the development of representation learning over the last few years for each modality and how they were combined to get a multimodal agent later.
Keyword: CLIP
Adap DP-FL: Differentially Private Federated Learning with Adaptive Noise
Authors: Jie Fu, Zhili Chen, Xiao Han
Subjects: Machine Learning (cs.LG); Cryptography and Security (cs.CR); Distributed, Parallel, and Cluster Computing (cs.DC)
Abstract
Federated learning seeks to address the issue of isolated data islands by making clients disclose only their local training models. However, it was demonstrated that private information could still be inferred by analyzing local model parameters, such as deep neural network model weights. Recently, differential privacy has been applied to federated learning to protect data privacy, but the noise added may degrade the learning performance much. Typically, in previous work, training parameters were clipped equally and noises were added uniformly. The heterogeneity and convergence of training parameters were simply not considered. In this paper, we propose a differentially private scheme for federated learning with adaptive noise (Adap DP-FL). Specifically, due to the gradient heterogeneity, we conduct adaptive gradient clipping for different clients and different rounds; due to the gradient convergence, we add decreasing noises accordingly. Extensive experiments on real-world datasets demonstrate that our Adap DP-FL outperforms previous methods significantly.
Model Extraction Attack against Self-supervised Speech Models
Authors: Tsu-Yuan Hsu, Chen-An Li, Tung-Yu Wu, Hung-yi Lee
Subjects: Sound (cs.SD); Computation and Language (cs.CL); Cryptography and Security (cs.CR); Machine Learning (cs.LG); Audio and Speech Processing (eess.AS)
Abstract
Self-supervised learning (SSL) speech models generate meaningful representations of given clips and achieve incredible performance across various downstream tasks. Model extraction attack (MEA) often refers to an adversary stealing the functionality of the victim model with only query access. In this work, we study the MEA problem against SSL speech model with a small number of queries. We propose a two-stage framework to extract the model. In the first stage, SSL is conducted on the large-scale unlabeled corpus to pre-train a small speech model. Secondly, we actively sample a small portion of clips from the unlabeled corpus and query the target model with these clips to acquire their representations as labels for the small model's second-stage training. Experiment results show that our sampling methods can effectively extract the target model without knowing any information about its model architecture.
Context-Aware Robust Fine-Tuning
Authors: Xiaofeng Mao, Yuefeng Chen, Xiaojun Jia, Rong Zhang, Hui Xue, Zhao Li
Subjects: Computer Vision and Pattern Recognition (cs.CV)
Abstract
Contrastive Language-Image Pre-trained (CLIP) models have zero-shot ability of classifying an image belonging to "[CLASS]" by using similarity between the image and the prompt sentence "a [CONTEXT] of [CLASS]". Based on exhaustive text cues in "[CONTEXT]", CLIP model is aware of different contexts, e.g. background, style, viewpoint, and exhibits unprecedented robustness against a wide range of distribution shifts. However, recent works find further fine-tuning of CLIP models improves accuracy but sacrifices the robustness on downstream tasks. We conduct an empirical investigation to show fine-tuning will corrupt the context-aware ability of pre-trained CLIP features. To solve this problem, we propose Context-Aware Robust Fine-tuning (CAR-FT). CAR-FT regularizes the model during fine-tuning to capture the context information. Specifically, we use zero-shot prompt weights to get the context distribution contained in the image. By minimizing the Kullback-Leibler Divergence (KLD) between context distributions induced by original/fine-tuned CLIP models, CAR-FT makes the context-aware ability of CLIP inherited into downstream tasks, and achieves both higher In-Distribution (ID) and Out-Of-Distribution (OOD) accuracy. The experimental results show CAR-FT achieves superior robustness on five OOD test datasets of ImageNet, and meanwhile brings accuracy gains on nine downstream tasks. Additionally, CAR-FT surpasses previous Domain Generalization (DG) methods and gets 78.5% averaged accuracy on DomainBed benchmark, building the new state-of-the-art.
SgVA-CLIP: Semantic-guided Visual Adapting of Vision-Language Models for Few-shot Image Classification
Authors: Fang Peng, Xiaoshan Yang, Changsheng Xu
Subjects: Computer Vision and Pattern Recognition (cs.CV); Multimedia (cs.MM)
Abstract
Although significant progress has been made in few-shot learning, most of existing few-shot learning methods require supervised pre-training on a large amount of samples of base classes, which limits their generalization ability in real world application. Recently, large-scale self-supervised vision-language models (e.g., CLIP) have provided a new paradigm for transferable visual representation learning. However, the pre-trained VLPs may neglect detailed visual information that is difficult to describe by language sentences, but important for learning an effective classifier in few-shot classification. To address the above problem, we propose a new framework, named Semantic-guided Visual Adapting (SgVA), which can effectively extend vision-language pre-trained models to produce discriminative task-specific visual features by comprehensively using a vision-specific contrastive loss, a cross-modal contrastive loss, and an implicit knowledge distillation. The implicit knowledge distillation is designed to transfer the fine-grained cross-modal knowledge to guide the updating of the vision adapter. State-of-the-art results on 13 datasets demonstrate that the adapted visual features can well complement the cross-modal features to improve few-shot image classification.
SuS-X: Training-Free Name-Only Transfer of Vision-Language Models
Authors: Vishaal Udandarao, Ankush Gupta, Samuel Albanie
Subjects: Computer Vision and Pattern Recognition (cs.CV); Computation and Language (cs.CL); Multimedia (cs.MM)
Abstract
Contrastive Language-Image Pre-training (CLIP) has emerged as a simple yet effective way to train large-scale vision-language models. CLIP demonstrates impressive zero-shot classification and retrieval on diverse downstream tasks. However, to leverage its full potential, fine-tuning still appears to be necessary. Fine-tuning the entire CLIP model can be resource-intensive and unstable. Moreover, recent methods that aim to circumvent this need for fine-tuning still require access to images from the target distribution. In this paper, we pursue a different approach and explore the regime of training-free "name-only transfer" in which the only knowledge we possess about the downstream task comprises the names of downstream target categories. We propose a novel method, SuS-X, consisting of two key building blocks -- SuS and TIP-X, that requires neither intensive fine-tuning nor costly labelled data. SuS-X achieves state-of-the-art zero-shot classification results on 19 benchmark datasets. We further show the utility of TIP-X in the training-free few-shot setting, where we again achieve state-of-the-art results over strong training-free baselines. Code is available at https://github.com/vishaal27/SuS-X.
DATID-3D: Diversity-Preserved Domain Adaptation Using Text-to-Image Diffusion for 3D Generative Model
Authors: Gwanghyun Kim, Se Young Chun
Subjects: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
Abstract
Recent 3D generative models have achieved remarkable performance in synthesizing high resolution photorealistic images with view consistency and detailed 3D shapes, but training them for diverse domains is challenging since it requires massive training images and their camera distribution information. Text-guided domain adaptation methods have shown impressive performance on converting the 2D generative model on one domain into the models on other domains with different styles by leveraging the CLIP (Contrastive Language-Image Pre-training), rather than collecting massive datasets for those domains. However, one drawback of them is that the sample diversity in the original generative model is not well-preserved in the domain-adapted generative models due to the deterministic nature of the CLIP text encoder. Text-guided domain adaptation will be even more challenging for 3D generative models not only because of catastrophic diversity loss, but also because of inferior text-image correspondence and poor image quality. Here we propose DATID-3D, a domain adaptation method tailored for 3D generative models using text-to-image diffusion models that can synthesize diverse images per text prompt without collecting additional images and camera information for the target domain. Unlike 3D extensions of prior text-guided domain adaptation methods, our novel pipeline was able to fine-tune the state-of-the-art 3D generator of the source domain to synthesize high resolution, multi-view consistent images in text-guided targeted domains without additional data, outperforming the existing text-guided domain adaptation methods in diversity and text-image correspondence. Furthermore, we propose and demonstrate diverse 3D image manipulations such as one-shot instance-selected adaptation and single-view manipulated 3D reconstruction to fully enjoy diversity in text.
NeuralLift-360: Lifting An In-the-wild 2D Photo to A 3D Object with 360° Views
Authors: Dejia Xu, Yifan Jiang, Peihao Wang, Zhiwen Fan, Yi Wang, Zhangyang Wang
Subjects: Computer Vision and Pattern Recognition (cs.CV)
Abstract
Virtual reality and augmented reality (XR) bring increasing demand for 3D content. However, creating high-quality 3D content requires tedious work that a human expert must do. In this work, we study the challenging task of lifting a single image to a 3D object and, for the first time, demonstrate the ability to generate a plausible 3D object with 360{\deg} views that correspond well with the given reference image. By conditioning on the reference image, our model can fulfill the everlasting curiosity for synthesizing novel views of objects from images. Our technique sheds light on a promising direction of easing the workflows for 3D artists and XR designers. We propose a novel framework, dubbed NeuralLift-360, that utilizes a depth-aware neural radiance representation (NeRF) and learns to craft the scene guided by denoising diffusion models. By introducing a ranking loss, our NeuralLift-360 can be guided with rough depth estimation in the wild. We also adopt a CLIP-guided sampling strategy for the diffusion prior to provide coherent guidance. Extensive experiments demonstrate that our NeuralLift-360 significantly outperforms existing state-of-the-art baselines. Project page: https://vita-group.github.io/NeuralLift-360/
Keyword: metric learning
From Forks to Forceps: A New Framework for Instance Segmentation of Surgical Instruments
Advancing Deep Metric Learning Through Multiple Batch Norms And Multi-Targeted Adversarial Examples
Intra-class Adaptive Augmentation with Neighbor Correction for Deep Metric Learning
Learning and Understanding a Disentangled Feature Representation for Hidden Parameters in Reinforcement Learning
Keyword: image retrieval
SLAN: Self-Locator Aided Network for Cross-Modal Understanding
Keyword: self-supervised
Handling Image and Label Resolution Mismatch in Remote Sensing
Survey on Self-Supervised Multimodal Representation Learning and Foundation Models
MoDA: Map style transfer for self-supervised Domain Adaptation of embodied agents
Textual Enhanced Contrastive Learning for Solving Math Word Problems
Model Extraction Attack against Self-supervised Speech Models
G-CMP: Graph-enhanced Contextual Matrix Profile for unsupervised anomaly detection in sensor-based remote health monitoring
SgVA-CLIP: Semantic-guided Visual Adapting of Vision-Language Models for Few-shot Image Classification
Exploring adaptation of VideoMAE for Audio-Visual Diarization & Social @ Ego4d Looking at me Challenge
On the power of foundation models
BARTSmiles: Generative Masked Language Models for Molecular Representations
Keyword: vision transformer
LUMix: Improving Mixup by Better Modelling Label Uncertainty
NoisyQuant: Noisy Bias-Enhanced Post-Training Activation Quantization for Vision Transformers
Metal-conscious Embedding for CBCT Projection Inpainting
Lightweight Structure-Aware Attention for Visual Understanding
RGB no more: Minimally-decoded JPEG Vision Transformers
Finding Differences Between Transformers and ConvNets Using Counterfactual Simulation Testing
Keyword: multimodal
Survey on Self-Supervised Multimodal Representation Learning and Foundation Models
Keyword: CLIP
Adap DP-FL: Differentially Private Federated Learning with Adaptive Noise
Model Extraction Attack against Self-supervised Speech Models
Context-Aware Robust Fine-Tuning
SgVA-CLIP: Semantic-guided Visual Adapting of Vision-Language Models for Few-shot Image Classification
SuS-X: Training-Free Name-Only Transfer of Vision-Language Models
DATID-3D: Diversity-Preserved Domain Adaptation Using Text-to-Image Diffusion for 3D Generative Model
NeuralLift-360: Lifting An In-the-wild 2D Photo to A 3D Object with 360° Views
Keyword: DALLE
There is no result