Anomaly detection is a fundamental yet challenging problem in machine learning due to the lack of label information. In this work, we propose a novel and powerful framework, dubbed as SLA$^2$P, for unsupervised anomaly detection. After extracting representative embeddings from raw data, we apply random projections to the features and regard features transformed by different projections as belonging to distinct pseudo classes. We then train a classifier network on these transformed features to perform self-supervised learning. Next we add adversarial perturbation to the transformed features to decrease their softmax scores of the predicted labels and design anomaly scores based on the predictive uncertainties of the classifier on these perturbed features. Our motivation is that because of the relatively small number and the decentralized modes of anomalies, 1) the pseudo label classifier's training concentrates more on learning the semantic information of normal data rather than anomalous data; 2) the transformed features of the normal data are more robust to the perturbations than those of the anomalies. Consequently, the perturbed transformed features of anomalies fail to be classified well and accordingly have lower anomaly scores than those of the normal samples. Extensive experiments on image, text and inherently tabular benchmark datasets back up our findings and indicate that SLA$^2$P achieves state-of-the-art results on unsupervised anomaly detection tasks consistently.
Personalized search plays a crucial role in improving user search experience owing to its ability to build user profiles based on historical behaviors. Previous studies have made great progress in extracting personal signals from the query log and learning user representations. However, neural personalized search is extremely dependent on sufficient data to train the user model. Data sparsity is an inevitable challenge for existing methods to learn high-quality user representations. Moreover, the overemphasis on final ranking quality leads to rough data representations and impairs the generalizability of the model. To tackle these issues, we propose a Personalized Search framework with Self-supervised Learning (PSSL) to enhance data representations. Specifically, we adopt a contrastive sampling method to extract paired self-supervised information from sequences of user behaviors in query logs. Four auxiliary tasks are designed to pre-train the sentence encoder and the sequence encoder used in the ranking model. They are optimized by contrastive loss which aims to close the distance between similar user sequences, queries, and documents. Experimental results on two datasets demonstrate that our proposed model PSSL achieves state-of-the-art performance compared with existing baselines.
Recent self-supervised models have demonstrated equal or better performance than supervised methods, opening for AI systems to learn visual representations from practically unlimited data. However, these methods are typically classification-based and thus ineffective for learning high-resolution feature maps that preserve precise spatial information. This work introduces superpixels to improve self-supervised learning of dense semantically rich visual concept embeddings. Decomposing images into a small set of visually coherent regions reduces the computational complexity by $\mathcal{O}(1000)$ while preserving detail. We experimentally show that contrasting over regions improves the effectiveness of contrastive learning methods, extends their applicability to high-resolution images, improves overclustering performance, superpixels are better than grids, and regional masking improves performance. The expressiveness of our dense embeddings is demonstrated by improving the SOTA unsupervised semantic segmentation benchmark on Cityscapes, and for convolutional models on COCO.
Transformer-based supervised pre-training achieves great performance in person re-identification (ReID). However, due to the domain gap between ImageNet and ReID datasets, it usually needs a larger pre-training dataset (e.g. ImageNet-21K) to boost the performance because of the strong data fitting ability of the transformer. To address this challenge, this work targets to mitigate the gap between the pre-training and ReID datasets from the perspective of data and model structure, respectively. We first investigate self-supervised learning (SSL) methods with Vision Transformer (ViT) pretrained on unlabelled person images (the LUPerson dataset), and empirically find it significantly surpasses ImageNet supervised pre-training models on ReID tasks. To further reduce the domain gap and accelerate the pre-training, the Catastrophic Forgetting Score (CFS) is proposed to evaluate the gap between pre-training and fine-tuning data. Based on CFS, a subset is selected via sampling relevant data close to the down-stream ReID data and filtering irrelevant data from the pre-training dataset. For the model structure, a ReID-specific module named IBN-based convolution stem (ICS) is proposed to bridge the domain gap by learning more invariant features. Extensive experiments have been conducted to fine-tune the pre-training models under supervised learning, unsupervised domain adaptation (UDA), and unsupervised learning (USL) settings. We successfully downscale the LUPerson dataset to 50% with no performance degradation. Finally, we achieve state-of-the-art performance on Market-1501 and MSMT17. For example, our ViT-S/16 achieves 91.3%/89.9%/89.6% mAP accuracy on Market1501 for supervised/UDA/USL ReID. Codes and models will be released to https://github.com/michuanhaohao/TransReID-SSL.
Music classification is a music information retrieval (MIR) task to classify music items to labels such as genre, mood, and instruments. It is also closely related to other concepts such as music similarity and musical preference. In this tutorial, we put our focus on two directions - the recent training schemes beyond supervised learning and the successful application of music classification models. The target audience for this web book is researchers and practitioners who are interested in state-of-the-art music classification research and building real-world applications. We assume the audience is familiar with the basic machine learning concepts. In this book, we present three lectures as follows: 1. Music classification overview: Task definition, applications, existing approaches, datasets, 2. Beyond supervised learning: Semi- and self-supervised learning for music classification, 3. Towards real-world applications: Less-discussed, yet important research issues in practice.
Self-supervised learning algorithms, including BERT and SimCLR, have enabled significant strides in fields like natural language processing, computer vision, and speech processing. However, these algorithms are domain-specific, meaning that new self-supervised learning algorithms must be developed for each new setting, including myriad healthcare, scientific, and multimodal domains. To catalyze progress toward domain-agnostic methods, we introduce DABS: a Domain-Agnostic Benchmark for Self-supervised learning. To perform well on DABS, an algorithm is evaluated on seven diverse domains: natural images, multichannel sensor data, English text, speech recordings, multilingual text, chest x-rays, and images with text descriptions. Each domain contains an unlabeled dataset for pretraining; the model is then is scored based on its downstream performance on a set of labeled tasks in the domain. We also present e-Mix and ShED: two baseline domain-agnostic algorithms; their relatively modest performance demonstrates that significant progress is needed before self-supervised learning is an out-of-the-box solution for arbitrary domains. Code for benchmark datasets and baseline algorithms is available at https://github.com/alextamkin/dabs.
Self-supervised learning is a powerful paradigm for representation learning on unlabelled images. A wealth of effective new methods based on instance matching rely on data-augmentation to drive learning, and these have reached a rough agreement on an augmentation scheme that optimises popular recognition benchmarks. However, there is strong reason to suspect that different tasks in computer vision require features to encode different (in)variances, and therefore likely require different augmentation strategies. In this paper, we measure the invariances learned by contrastive methods and confirm that they do learn invariance to the augmentations used and further show that this invariance largely transfers to related real-world changes in pose and lighting. We show that learned invariances strongly affect downstream task performance and confirm that different downstream tasks benefit from polar opposite (in)variances, leading to performance loss when the standard augmentation strategy is used. Finally, we demonstrate that a simple fusion of representations with complementary invariances ensures wide transferability to all the diverse downstream tasks considered.
Despite the large progress in supervised learning with neural networks, there are significant challenges in obtaining high-quality, large-scale and accurately labelled datasets. In such a context, how to learn in the presence of noisy labels has received more and more attention. As a relatively complex problem, in order to achieve good results, current approaches often integrate components from several fields, such as supervised learning, semi-supervised learning, transfer learning and resulting in complicated methods. Furthermore, they often make multiple assumptions about the type of noise of the data. This affects the model robustness and limits its performance under different noise conditions. In this paper, we consider a novel problem setting, Learning with Unknown Label Noise}(LULN), that is, learning when both the degree and the type of noise are unknown. Under this setting, unlike previous methods that often introduce multiple assumptions and lead to complex solutions, we propose a simple, efficient and robust framework named Sample Selection and Relabelling(SSR), that with a minimal number of hyperparameters achieves SOTA results in various conditions. At the heart of our method is a sample selection and relabelling mechanism based on a non-parametric KNN classifier~(NPK) $g_q$ and a parametric model classifier~(PMC) $g_p$, respectively, to select the clean samples and gradually relabel the noisy samples. Without bells and whistles, such as model co-training, self-supervised pre-training and semi-supervised learning, and with robustness concerning the settings of its few hyper-parameters, our method significantly surpasses previous methods on both CIFAR10/CIFAR100 with synthetic noise and real-world noisy datasets such as WebVision, Clothing1M and ANIMAL-10N. Code is available at https://github.com/MrChenFeng/SSR_BMVC2022.
We introduce PyTorchVideo, an open-source deep-learning library that provides a rich set of modular, efficient, and reproducible components for a variety of video understanding tasks, including classification, detection, self-supervised learning, and low-level processing. The library covers a full stack of video understanding tools including multimodal data loading, transformations, and models that reproduce state-of-the-art performance. PyTorchVideo further supports hardware acceleration that enables real-time inference on mobile devices. The library is based on PyTorch and can be used by any training framework; for example, PyTorchLightning, PySlowFast, or Classy Vision. PyTorchVideo is available at https://pytorchvideo.org/
Reconstructing high-fidelity 3D facial texture from a single image is a quite challenging task due to the lack of complete face information and the domain gap between the 3D face and 2D image. Further, obtaining re-renderable 3D faces has become a strongly desired property in many applications, where the term 're-renderable' demands the facial texture to be spatially complete and disentangled with environmental illumination. In this paper, we propose a new self-supervised deep learning framework for reconstructing high-quality and re-renderable facial albedos from single-view images in-the-wild. Our main idea is to first utilize a prior generation module based on the 3DMM proxy model to produce an unwrapped texture and a globally parameterized prior albedo. Then we apply a detail refinement module to synthesize the final texture with both high-frequency details and completeness. To further make facial textures disentangled with illumination, we propose a novel detailed illumination representation which is reconstructed with the detailed albedo together. We also design several novel regularization losses on both the albedo and illumination maps to facilitate the disentanglement of these two factors. Finally, by leveraging a differentiable renderer, each face attribute can be jointly trained in a self-supervised manner without requiring ground-truth facial reflectance. Extensive comparisons and ablation studies on challenging datasets demonstrate that our framework outperforms state-of-the-art approaches.
Clustering of hyperspectral images is a fundamental but challenging task. The recent development of hyperspectral image clustering has evolved from shallow models to deep and achieved promising results in many benchmark datasets. However, their poor scalability, robustness, and generalization ability, mainly resulting from their offline clustering scenarios, greatly limit their application to large-scale hyperspectral data. To circumvent these problems, we present a scalable deep online clustering model, named Spectral-Spatial Contrastive Clustering (SSCC), based on self-supervised learning. Specifically, we exploit a symmetric twin neural network comprised of a projection head with a dimensionality of the cluster number to conduct dual contrastive learning from a spectral-spatial augmentation pool. We define the objective function by implicitly encouraging within-cluster similarity and reducing between-cluster redundancy. The resulting approach is trained in an end-to-end fashion by batch-wise optimization, making it robust in large-scale data and resulting in good generalization ability for unseen data. Extensive experiments on three hyperspectral image benchmarks demonstrate the effectiveness of our approach and show that we advance the state-of-the-art approaches by large margins.
Humans can naturally and effectively find salient regions in complex scenes. Motivated by this observation, attention mechanisms were introduced into computer vision with the aim of imitating this aspect of the human visual system. Such an attention mechanism can be regarded as a dynamic weight adjustment process based on features of the input image. Attention mechanisms have achieved great success in many visual tasks, including image classification, object detection, semantic segmentation, video understanding, image generation, 3D vision, multi-modal tasks and self-supervised learning. In this survey, we provide a comprehensive review of various attention mechanisms in computer vision and categorize them according to approach, such as channel attention, spatial attention, temporal attention and branch attention; a related repository https://github.com/MenghaoGuo/Awesome-Vision-Attentions is dedicated to collecting related work. We also suggest future directions for attention mechanism research.
Object detection has achieved promising performance on clean datasets, but how to achieve better tradeoff between the adversarial robustness and clean precision is still under-explored. Adversarial training is the mainstream method to improve robustness, but most of the works will sacrifice clean precision to gain robustness than standard training. In this paper, we propose Unified Decoupled Feature Alignment (UDFA), a novel fine-tuning paradigm which achieves better performance than existing methods, by fully exploring the combination between self-knowledge distillation and adversarial training for object detection. We first use decoupled fore/back-ground features to construct self-knowledge distillation branch between clean feature representation from pretrained detector (served as teacher) and adversarial feature representation from student detector. Then we explore the self-knowledge distillation from a new angle by decoupling original branch into a self-supervised learning branch and a new self-knowledge distillation branch. With extensive experiments on the PASCAL-VOC and MS-COCO benchmarks, the evaluation results show that UDFA can surpass the standard training and state-of-the-art adversarial training methods for object detection. For example, compared with teacher detector, our approach on GFLV2 with ResNet-50 improves clean precision by 2.2 AP on PASCAL-VOC; compared with SOTA adversarial training methods, our approach improves clean precision by 1.6 AP, while improving adversarial robustness by 0.5 AP. Our code will be available at https://github.com/grispeut/udfa.
In this paper, we propose a self-supervised learning procedure for training a robust multi-object tracking (MOT) model given only unlabeled video. While several self-supervisory learning signals have been proposed in prior work on single-object tracking, such as color propagation and cycle-consistency, these signals cannot be directly applied for training RNN models, which are needed to achieve accurate MOT: they yield degenerate models that, for instance, always match new detections to tracks with the closest initial detections. We propose a novel self-supervisory signal that we call cross-input consistency: we construct two distinct inputs for the same sequence of video, by hiding different information about the sequence in each input. We then compute tracks in that sequence by applying an RNN model independently on each input, and train the model to produce consistent tracks across the two inputs. We evaluate our unsupervised method on MOT17 and KITTI -- remarkably, we find that, despite training only on unlabeled video, our unsupervised approach outperforms four supervised methods published in the last 1--2 years, including Tracktor++, FAMNet, GSM, and mmMOT.
This paper shows that masked autoencoders (MAE) are scalable self-supervised learners for computer vision. Our MAE approach is simple: we mask random patches of the input image and reconstruct the missing pixels. It is based on two core designs. First, we develop an asymmetric encoder-decoder architecture, with an encoder that operates only on the visible subset of patches (without mask tokens), along with a lightweight decoder that reconstructs the original image from the latent representation and mask tokens. Second, we find that masking a high proportion of the input image, e.g., 75%, yields a nontrivial and meaningful self-supervisory task. Coupling these two designs enables us to train large models efficiently and effectively: we accelerate training (by 3x or more) and improve accuracy. Our scalable approach allows for learning high-capacity models that generalize well: e.g., a vanilla ViT-Huge model achieves the best accuracy (87.8%) among methods that use only ImageNet-1K data. Transfer performance in downstream tasks outperforms supervised pre-training and shows promising scaling behavior.
We present CrissCross, a self-supervised framework for learning audio-visual representations. A novel notion is introduced in our framework whereby in addition to learning the intra-modal and standard 'synchronous' cross-modal relations, CrissCross also learns 'asynchronous' cross-modal relationships. We perform in-depth studies showing that by relaxing the temporal synchronicity between the audio and visual modalities, the network learns strong generalized representations useful for a variety of downstream tasks. To pretrain our proposed solution, we use 3 different datasets with varying sizes, Kinetics-Sound, Kinetics400, and AudioSet. The learned representations are evaluated on a number of downstream tasks namely action recognition, sound classification, and action retrieval. Our experiments show that CrissCross either outperforms or achieves performances on par with the current state-of-the-art self-supervised methods on action recognition and action retrieval with UCF101 and HMDB51, as well as sound classification with ESC50 and DCASE. Moreover, CrissCross outperforms fully-supervised pretraining while pretrained on Kinetics-Sound. The codes and pretrained models are available on the project website.
Recently, adapting the idea of self-supervised learning (SSL) on continuous speech has started gaining attention. SSL models pre-trained on a huge amount of unlabeled audio can generate general-purpose representations that benefit a wide variety of speech processing tasks. Despite their ubiquitous deployment, however, the potential privacy risks of these models have not been well investigated. In this paper, we present the first privacy analysis on several SSL speech models using Membership Inference Attacks (MIA) under black-box access. The experiment results show that these pre-trained models are vulnerable to MIA and prone to membership information leakage with high Area Under the Curve (AUC) in both utterance-level and speaker-level. Furthermore, we also conduct several ablation studies to understand the factors that contribute to the success of MIA.
Due to the scarcity of labeled data, using supervised models pre-trained on ImageNet is a de facto standard in remote sensing scene classification. Recently, the availability of larger high resolution remote sensing (HRRS) image datasets and progress in self-supervised learning have brought up the questions of whether supervised ImageNet pre-training is still necessary for remote sensing scene classification and would supervised pre-training on HRRS image datasets or self-supervised pre-training on ImageNet achieve better results on target remote sensing scene classification tasks. To answer these questions, in this paper we both train models from scratch and fine-tune supervised and self-supervised ImageNet models on several HRRS image datasets. We also evaluate the transferability of learned representations to HRRS scene classification tasks and show that self-supervised pre-training outperforms the supervised one, while the performance of HRRS pre-training is similar to self-supervised pre-training or slightly lower. Finally, we propose using an ImageNet pre-trained model combined with a second round of pre-training using in-domain HRRS images, i.e. domain-adaptive pre-training. The experimental results show that domain-adaptive pre-training results in models that achieve state-of-the-art results on HRRS scene classification benchmarks. The source code and pre-trained models are available at \url{https://github.com/risojevicv/RSSC-transfer}.
Pre-training lays the foundation for recent successes in radiograph analysis supported by deep learning. It learns transferable image representations by conducting large-scale fully-supervised or self-supervised learning on a source domain. However, supervised pre-training requires a complex and labor intensive two-stage human-assisted annotation process while self-supervised learning cannot compete with the supervised paradigm. To tackle these issues, we propose a cross-supervised methodology named REviewing FreE-text Reports for Supervision (REFERS), which acquires free supervision signals from original radiology reports accompanying the radiographs. The proposed approach employs a vision transformer and is designed to learn joint representations from multiple views within every patient study. REFERS outperforms its transfer learning and self-supervised learning counterparts on 4 well-known X-ray datasets under extremely limited supervision. Moreover, REFERS even surpasses methods based on a source domain of radiographs with human-assisted structured labels. Thus REFERS has the potential to replace canonical pre-training methodologies.
Unsupervised graph representation learning is a non-trivial topic. The success of contrastive methods in the unsupervised representation learning on structured data inspires similar attempts on the graph. Existing graph contrastive learning (GCL) aims to learn the invariance across multiple augmentation views, which renders it heavily reliant on the handcrafted graph augmentations. However, inappropriate graph data augmentations can potentially jeopardize such invariance. In this paper, we show the potential hazards of inappropriate augmentations and then propose a novel Collaborative Graph Contrastive Learning framework (CGCL). This framework harnesses multiple graph encoders to observe the graph. Features observed from different encoders serve as the contrastive views in contrastive learning, which avoids inducing unstable perturbation and guarantees the invariance. To ensure the collaboration among diverse graph encoders, we propose the concepts of asymmetric architecture and complementary encoders as the design principle. To further prove the rationality, we utilize two quantitative metrics to measure the assembly of CGCL respectively. Extensive experiments demonstrate the advantages of CGCL in unsupervised graph-level representation learning and the potential of collaborative framework. The source code for reproducibility is available at https://github.com/zhangtia16/CGCL