PyTorchKR Papers

최신 논문

총 5,358편
태그·기간 등 상세 필터

카테고리 탐색

로그인 후 이 태그 구독

arXiv:2203.17205공식 코드 · Papers with Code

Leverage Your Local and Global Representations: A New Self-Supervised Learning Strategy

Self-supervised learning (SSL) methods aim to learn view-invariant representations by maximizing the similarity between the features extracted from different crops of the same image regardless of cropping size and content. In essence, this strategy ignores the fact that two crops may truly contain different image information, e.g., background and small objects, and thus tends to restrain the diversity of the learned representations. In this work, we address this issue by introducing a new self-supervised learning strategy, LoGo, that explicitly reasons about Local and Global crops. To achieve view invariance, LoGo encourages similarity between global crops from the same image, as well as between a global and a local crop. However, to correctly encode the fact that the content of smaller crops may differ entirely, LoGo promotes two local crops to have dissimilar representations, while being close to global crops. Our LoGo strategy can easily be applied to existing SSL methods. Our extensive experiments on a variety of datasets and using different self-supervised learning frameworks validate its superiority over existing approaches. Noticeably, we achieve better results than supervised models on transfer learning when using only 1/10 of the data.

arXiv:2203.16983

Self-distillation Augmented Masked Autoencoders for Histopathological Image Classification

Self-supervised learning (SSL) has drawn increasing attention in histopathological image analysis in recent years. Compared to contrastive learning which is troubled with the false negative problem, i.e., semantically similar images are selected as negative samples, masked autoencoders (MAE) building SSL from a generative paradigm is probably a more appropriate pre-training. In this paper, we introduce MAE and verify the effect of visible patches for histopathological image understanding. Moreover, a novel SD-MAE model is proposed to enable a self-distillation augmented MAE. Besides the reconstruction loss on masked image patches, SD-MAE further imposes the self-distillation loss on visible patches to enhance the representational capacity of the encoder located shallow layer. We apply SD-MAE to histopathological image classification, cell segmentation and object detection. Experiments demonstrate that SD-MAE shows highly competitive performance when compared with other SSL methods in these tasks.

arXiv:2203.15937공식 코드 · Papers with Code

Improving Mispronunciation Detection with Wav2vec2-based Momentum Pseudo-Labeling for Accentedness and Intelligibility Assessment

Current leading mispronunciation detection and diagnosis (MDD) systems achieve promising performance via end-to-end phoneme recognition. One challenge of such end-to-end solutions is the scarcity of human-annotated phonemes on natural L2 speech. In this work, we leverage unlabeled L2 speech via a pseudo-labeling (PL) procedure and extend the fine-tuning approach based on pre-trained self-supervised learning (SSL) models. Specifically, we use Wav2vec 2.0 as our SSL model, and fine-tune it using original labeled L2 speech samples plus the created pseudo-labeled L2 speech samples. Our pseudo labels are dynamic and are produced by an ensemble of the online model on-the-fly, which ensures that our model is robust to pseudo label noise. We show that fine-tuning with pseudo labels achieves a 5.35% phoneme error rate reduction and 2.48% MDD F1 score improvement over a labeled-samples-only fine-tuning baseline. The proposed PL method is also shown to outperform conventional offline PL methods. Compared to the state-of-the-art MDD systems, our MDD solution produces a more accurate and consistent phonetic error diagnosis. In addition, we conduct an open test on a separate UTD-4Accents dataset, where our system recognition outputs show a strong correlation with human perception, based on accentedness and intelligibility.

arXiv:2203.15876공식 코드 · Papers with Code

Self-Supervised Learning for Recommender Systems: A Survey

In recent years, neural architecture-based recommender systems have achieved tremendous success, but they still fall short of expectation when dealing with highly sparse data. Self-supervised learning (SSL), as an emerging technique for learning from unlabeled data, has attracted considerable attention as a potential solution to this issue. This survey paper presents a systematic and timely review of research efforts on self-supervised recommendation (SSR). Specifically, we propose an exclusive definition of SSR, on top of which we develop a comprehensive taxonomy to divide existing SSR methods into four categories: contrastive, generative, predictive, and hybrid. For each category, we elucidate its concept and formulation, the involved methods, as well as its pros and cons. Furthermore, to facilitate empirical comparison, we release an open-source library SELFRec (https://github.com/Coder-Yu/SELFRec), which incorporates a wide range of SSR models and benchmark datasets. Through rigorous experiments using this library, we derive and report some significant findings regarding the selection of self-supervised signals for enhancing recommendation. Finally, we shed light on the limitations in the current research and outline the future research directions.

arXiv:2203.15943공식 코드 · Papers with Code

Self-Supervised Leaf Segmentation under Complex Lighting Conditions

As an essential prerequisite task in image-based plant phenotyping, leaf segmentation has garnered increasing attention in recent years. While self-supervised learning is emerging as an effective alternative to various computer vision tasks, its adaptation for image-based plant phenotyping remains rather unexplored. In this work, we present a self-supervised leaf segmentation framework consisting of a self-supervised semantic segmentation model, a color-based leaf segmentation algorithm, and a self-supervised color correction model. The self-supervised semantic segmentation model groups the semantically similar pixels by iteratively referring to the self-contained information, allowing the pixels of the same semantic object to be jointly considered by the color-based leaf segmentation algorithm for identifying the leaf regions. Additionally, we propose to use a self-supervised color correction model for images taken under complex illumination conditions. Experimental results on datasets of different plant species demonstrate the potential of the proposed self-supervised framework in achieving effective and generalizable leaf segmentation.

arXiv:2203.16262

How Does SimSiam Avoid Collapse Without Negative Samples? A Unified Understanding with Self-supervised Contrastive Learning

To avoid collapse in self-supervised learning (SSL), a contrastive loss is widely used but often requires a large number of negative samples. Without negative samples yet achieving competitive performance, a recent work has attracted significant attention for providing a minimalist simple Siamese (SimSiam) method to avoid collapse. However, the reason for how it avoids collapse without negative samples remains not fully clear and our investigation starts by revisiting the explanatory claims in the original SimSiam. After refuting their claims, we introduce vector decomposition for analyzing the collapse based on the gradient analysis of the $l_2$-normalized representation vector. This yields a unified perspective on how negative samples and SimSiam alleviate collapse. Such a unified perspective comes timely for understanding the recent progress in SSL.

arXiv:2203.15537공식 코드 · Papers with Code

On Metric Learning for Audio-Text Cross-Modal Retrieval

Audio-text retrieval aims at retrieving a target audio clip or caption from a pool of candidates given a query in another modality. Solving such cross-modal retrieval task is challenging because it not only requires learning robust feature representations for both modalities, but also requires capturing the fine-grained alignment between these two modalities. Existing cross-modal retrieval models are mostly optimized by metric learning objectives as both of them attempt to map data to an embedding space, where similar data are close together and dissimilar data are far apart. Unlike other cross-modal retrieval tasks such as image-text and video-text retrievals, audio-text retrieval is still an unexplored task. In this work, we aim to study the impact of different metric learning objectives on the audio-text retrieval task. We present an extensive evaluation of popular metric learning objectives on the AudioCaps and Clotho datasets. We demonstrate that NT-Xent loss adapted from self-supervised learning shows stable performance across different datasets and training settings, and outperforms the popular triplet-based losses. Our code is available at https://github.com/XinhaoMei/audio-text_retrieval.

arXiv:2203.15519

Learning neural audio features without supervision

Deep audio classification, traditionally cast as training a deep neural network on top of mel-filterbanks in a supervised fashion, has recently benefited from two independent lines of work. The first one explores "learnable frontends", i.e., neural modules that produce a learnable time-frequency representation, to overcome limitations of fixed features. The second one uses self-supervised learning to leverage unprecedented scales of pre-training data. In this work, we study the feasibility of combining both approaches, i.e., pre-training learnable frontend jointly with the main architecture for downstream classification. First, we show that pretraining two previously proposed frontends (SincNet and LEAF) on Audioset drastically improves linear-probe performance over fixed mel-filterbanks, suggesting that learnable time-frequency representations can benefit self-supervised pre-training even more than supervised training. Surprisingly, randomly initialized learnable filterbanks outperform mel-scaled initialization in the self-supervised setting, a counter-intuitive result that questions the appropriateness of strong priors when designing learnable filters. Through exploratory analysis of the learned frontend components, we uncover crucial differences in properties of these frontends when used in a supervised and self-supervised setting, especially the affinity of self-supervised filters to diverge significantly from the mel-scale to model a broader range of frequencies.

arXiv:2203.15089

Learning Optical Flow, Depth, and Scene Flow without Real-World Labels

Self-supervised monocular depth estimation enables robots to learn 3D perception from raw video streams. This scalable approach leverages projective geometry and ego-motion to learn via view synthesis, assuming the world is mostly static. Dynamic scenes, which are common in autonomous driving and human-robot interaction, violate this assumption. Therefore, they require modeling dynamic objects explicitly, for instance via estimating pixel-wise 3D motion, i.e. scene flow. However, the simultaneous self-supervised learning of depth and scene flow is ill-posed, as there are infinitely many combinations that result in the same 3D point. In this paper we propose DRAFT, a new method capable of jointly learning depth, optical flow, and scene flow by combining synthetic data with geometric self-supervision. Building upon the RAFT architecture, we learn optical flow as an intermediate task to bootstrap depth and scene flow learning via triangulation. Our algorithm also leverages temporal and geometric consistency losses across tasks to improve multi-task learning. Our DRAFT architecture simultaneously establishes a new state of the art in all three tasks in the self-supervised monocular setting on the standard KITTI benchmark. Project page: https://sites.google.com/tri.global/draft.

arXiv:2203.15205공식 코드 · Papers with Code

SPAct: Self-supervised Privacy Preservation for Action Recognition

Visual private information leakage is an emerging key issue for the fast growing applications of video understanding like activity recognition. Existing approaches for mitigating privacy leakage in action recognition require privacy labels along with the action labels from the video dataset. However, annotating frames of video dataset for privacy labels is not feasible. Recent developments of self-supervised learning (SSL) have unleashed the untapped potential of the unlabeled data. For the first time, we present a novel training framework which removes privacy information from input video in a self-supervised manner without requiring privacy labels. Our training framework consists of three main components: anonymization function, self-supervised privacy removal branch, and action recognition branch. We train our framework using a minimax optimization strategy to minimize the action recognition cost function and maximize the privacy cost function through a contrastive self-supervised loss. Employing existing protocols of known-action and privacy attributes, our framework achieves a competitive action-privacy trade-off to the existing state-of-the-art supervised methods. In addition, we introduce a new protocol to evaluate the generalization of learned the anonymization function to novel-action and privacy attributes and show that our self-supervised framework outperforms existing supervised methods. Code available at: https://github.com/DAVEISHAN/SPAct

arXiv:2203.15171

Self-Supervised Light Field Depth Estimation Using Epipolar Plane Images

Exploiting light field data makes it possible to obtain dense and accurate depth map. However, synthetic scenes with limited disparity range cannot contain the diversity of real scenes. By training in synthetic data, current learning-based methods do not perform well in real scenes. In this paper, we propose a self-supervised learning framework for light field depth estimation. Different from the existing end-to-end training methods using disparity label per pixel, our approach implements network training by estimating EPI disparity shift after refocusing, which extends the disparity range of epipolar lines. To reduce the sensitivity of EPI to noise, we propose a new input mode called EPI-Stack, which stacks EPIs in the view dimension. This method is less sensitive to noise scenes than traditional input mode and improves the efficiency of estimation. Compared with other state-of-the-art methods, the proposed method can also obtain higher quality results in real-world scenarios, especially in the complex occlusion and depth discontinuity.

arXiv:2203.15508공식 코드 · Papers with Code

Improving Contrastive Learning with Model Augmentation

The sequential recommendation aims at predicting the next items in user behaviors, which can be solved by characterizing item relationships in sequences. Due to the data sparsity and noise issues in sequences, a new self-supervised learning (SSL) paradigm is proposed to improve the performance, which employs contrastive learning between positive and negative views of sequences. However, existing methods all construct views by adopting augmentation from data perspectives, while we argue that 1) optimal data augmentation methods are hard to devise, 2) data augmentation methods destroy sequential correlations, and 3) data augmentation fails to incorporate comprehensive self-supervised signals. Therefore, we investigate the possibility of model augmentation to construct view pairs. We propose three levels of model augmentation methods: neuron masking, layer dropping, and encoder complementing. This work opens up a novel direction in constructing views for contrastive SSL. Experiments verify the efficacy of model augmentation for the SSL in the sequential recommendation. Code is available\footnote{\url{https://github.com/salesforce/SRMA}}.

arXiv:2203.14553공식 코드 · Papers with Code

Investigating Active-learning-based Training Data Selection for Speech Spoofing Countermeasure

Training a spoofing countermeasure (CM) that generalizes to various unseen data is desired but challenging. While methods such as data augmentation and self-supervised learning are applicable, the imperfect CM performance on diverse test sets still calls for additional strategies. This study took the initiative and investigated CM training using active learning (AL), a framework that iteratively selects useful data from a large pool set and fine-tunes the CM. This study compared a few methods to measure the data usefulness and the impact of using different pool sets collected from various sources. The results showed that the AL-based CMs achieved better generalization than our strong baseline on multiple test tests. Furthermore, compared with a top-line CM that simply used the whole data pool set for training, the AL-based CMs achieved similar performance using less training data. Although no single best configuration was found for AL, the rule of thumb is to include diverse spoof and bona fide data in the pool set and to avoid any AL data selection method that selects the data that the CM feels confident in.

arXiv:2203.14525

Curriculum learning for self-supervised speaker verification

The goal of this paper is to train effective self-supervised speaker representations without identity labels. We propose two curriculum learning strategies within a self-supervised learning framework. The first strategy aims to gradually increase the number of speakers in the training phase by enlarging the used portion of the train dataset. The second strategy applies various data augmentations to more utterances within a mini-batch as the training proceeds. A range of experiments conducted using the DINO self-supervised framework on the VoxCeleb1 evaluation protocol demonstrates the effectiveness of our proposed curriculum learning strategies. We report a competitive equal error rate of 4.47% with a single-phase training, and we also demonstrate that the performance further improves to 1.84% by fine-tuning on a small labelled dataset.

arXiv:2203.14044공식 코드 · Papers with Code

Contrastive Graph Learning for Population-based fMRI Classification

Contrastive self-supervised learning has recently benefited fMRI classification with inductive biases. Its weak label reliance prevents overfitting on small medical datasets and tackles the high intraclass variances. Nonetheless, existing contrastive methods generate resemblant pairs only on pixel-level features of 3D medical images, while the functional connectivity that reveals critical cognitive information is under-explored. Additionally, existing methods predict labels on individual contrastive representation without recognizing neighbouring information in the patient group, whereas interpatient contrast can act as a similarity measure suitable for population-based classification. We hereby proposed contrastive functional connectivity graph learning for population-based fMRI classification. Representations on the functional connectivity graphs are "repelled" for heterogeneous patient pairs meanwhile homogeneous pairs "attract" each other. Then a dynamic population graph that strengthens the connections between similar patients is updated for classification. Experiments on a multi-site dataset ADHD200 validate the superiority of the proposed method on various metrics. We initially visualize the population relationships and exploit potential subtypes.

arXiv:2203.13865공식 코드 · Papers with Code

Intelligent Masking: Deep Q-Learning for Context Encoding in Medical Image Analysis

The need for a large amount of labeled data in the supervised setting has led recent studies to utilize self-supervised learning to pre-train deep neural networks using unlabeled data. Many self-supervised training strategies have been investigated especially for medical datasets to leverage the information available in the much fewer unlabeled data. One of the fundamental strategies in image-based self-supervision is context prediction. In this approach, a model is trained to reconstruct the contents of an arbitrary missing region of an image based on its surroundings. However, the existing methods adopt a random and blind masking approach by focusing uniformly on all regions of the images. This approach results in a lot of unnecessary network updates that cause the model to forget the rich extracted features. In this work, we develop a novel self-supervised approach that occludes targeted regions to improve the pre-training procedure. To this end, we propose a reinforcement learning-based agent which learns to intelligently mask input images through deep Q-learning. We show that training the agent against the prediction model can significantly improve the semantic features extracted for downstream classification tasks. We perform our experiments on two public datasets for diagnosing breast cancer in the ultrasound images and detecting lower-grade glioma with MR images. In our experiments, we show that our novel masking strategy advances the learned features according to the performance on the classification task in terms of accuracy, macro F1, and AUROC.

arXiv:2203.14581

S2-Net: Self-supervision Guided Feature Representation Learning for Cross-Modality Images

Combining the respective advantages of cross-modality images can compensate for the lack of information in the single modality, which has attracted increasing attention of researchers into multi-modal image matching tasks. Meanwhile, due to the great appearance differences between cross-modality image pairs, it often fails to make the feature representations of correspondences as close as possible. In this letter, we design a cross-modality feature representation learning network, S2-Net, which is based on the recently successful detect-and-describe pipeline, originally proposed for visible images but adapted to work with cross-modality image pairs. To solve the consequent problem of optimization difficulties, we introduce self-supervised learning with a well-designed loss function to guide the training without discarding the original advantages. This novel strategy simulates image pairs in the same modality, which is also a useful guide for the training of cross-modality images. Notably, it does not require additional data but significantly improves the performance and is even workable for all methods of the detect-and-describe pipeline. Extensive experiments are conducted to evaluate the performance of the strategy we proposed, compared to both handcrafted and deep learning-based methods. Results show that our elegant formulation of combined optimization of supervised and self-supervised learning outperforms state-of-the-arts on RoadScene and RGB-NIR datasets.

arXiv:2203.14240공식 코드 · Papers with Code

Audio-Adaptive Activity Recognition Across Video Domains

This paper strives for activity recognition under domain shift, for example caused by change of scenery or camera viewpoint. The leading approaches reduce the shift in activity appearance by adversarial training and self-supervised learning. Different from these vision-focused works we leverage activity sounds for domain adaptation as they have less variance across domains and can reliably indicate which activities are not happening. We propose an audio-adaptive encoder and associated learning methods that discriminatively adjust the visual feature representation as well as addressing shifts in the semantic distribution. To further eliminate domain-specific features and include domain-invariant activity sounds for recognition, an audio-infused recognizer is proposed, which effectively models the cross-modal interaction across domains. We also introduce the new task of actor shift, with a corresponding audio-visual dataset, to challenge our method with situations where the activity appearance changes dramatically. Experiments on this dataset, EPIC-Kitchens and CharadesEgo show the effectiveness of our approach.

arXiv:2203.14221공식 코드 · Papers with Code

How Severe is Benchmark-Sensitivity in Video Self-Supervised Learning?

Despite the recent success of video self-supervised learning models, there is much still to be understood about their generalization capability. In this paper, we investigate how sensitive video self-supervised learning is to the current conventional benchmark and whether methods generalize beyond the canonical evaluation setting. We do this across four different factors of sensitivity: domain, samples, actions and task. Our study which encompasses over 500 experiments on 7 video datasets, 9 self-supervised methods and 6 video understanding tasks, reveals that current benchmarks in video self-supervised learning are not good indicators of generalization along these sensitivity factors. Further, we find that self-supervised methods considerably lag behind vanilla supervised pre-training, especially when domain shift is large and the amount of available downstream samples are low. From our analysis, we distill the SEVERE-benchmark, a subset of our experiments, and discuss its implication for evaluating the generalizability of representations obtained by existing and future self-supervised video learning methods.

arXiv:2203.14084공식 코드 · Papers with Code

3D-OAE: Occlusion Auto-Encoders for Self-Supervised Learning on Point Clouds

The manual annotation for large-scale point clouds is still tedious and unavailable for many harsh real-world tasks. Self-supervised learning, which is used on raw and unlabeled data to pre-train deep neural networks, is a promising approach to address this issue. Existing works usually take the common aid from auto-encoders to establish the self-supervision by the self-reconstruction schema. However, the previous auto-encoders merely focus on the global shapes and do not distinguish the local and global geometric features apart. To address this problem, we present a novel and efficient self-supervised point cloud representation learning framework, named 3D Occlusion Auto-Encoder (3D-OAE), to facilitate the detailed supervision inherited in local regions and global shapes. We propose to randomly occlude some local patches of point clouds and establish the supervision via inpainting the occluded patches using the remaining ones. Specifically, we design an asymmetrical encoder-decoder architecture based on standard Transformer, where the encoder operates only on the visible subset of patches to learn local patterns, and a lightweight decoder is designed to leverage these visible patterns to infer the missing geometries via self-attention. We find that occluding a very high proportion of the input point cloud (e.g. 75%) will still yield a nontrivial self-supervisory performance, which enables us to achieve 3-4 times faster during training but also improve accuracy. Experimental results show that our approach outperforms the state-of-the-art on a diverse range of downstream discriminative and generative tasks.