Local representation learning has been a key challenge to promote the performance of the histopathological whole slide images analysis. The previous representation learning methods followed the supervised learning paradigm. However, manual annotation for large-scale WSIs is time-consuming and labor-intensive. Hence, the self-supervised contrastive learning has recently attracted intensive attention. The present contrastive learning methods treat each sample as a single class, which suffers from class collision problems, especially in the domain of histopathology image analysis. In this paper, we proposed a novel contrastive representation learning framework named Lesion-Aware Contrastive Learning (LACL) for histopathology whole slide image analysis. We built a lesion queue based on the memory bank structure to store the representations of different classes of WSIs, which allowed the contrastive model to selectively define the negative pairs during the training. Moreover, We designed a queue refinement strategy to purify the representations stored in the lesion queue. The experimental results demonstrate that LACL achieves the best performance in histopathology image representation learning on different datasets, and outperforms state-of-the-art methods under different WSI classification benchmarks. The code is available at https://github.com/junl21/lacl.
Vision Transformer (ViT) has shown its advantages over the convolutional neural network (CNN) with its ability to capture global long-range dependencies for visual representation learning. Besides ViT, contrastive learning is another popular research topic recently. While previous contrastive learning works are mostly based on CNNs, some recent studies have attempted to combine ViT and contrastive learning for enhanced self-supervised learning. Despite the considerable progress, these combinations of ViT and contrastive learning mostly focus on the instance-level contrastiveness, which often overlook the global contrastiveness and also lack the ability to directly learn the clustering result (e.g., for images). In view of this, this paper presents a novel deep clustering approach termed Vision Transformer for Contrastive Clustering (VTCC), which for the first time, to our knowledge, unifies the Transformer and the contrastive learning for the image clustering task. Specifically, with two random augmentations performed on each image, we utilize a ViT encoder with two weight-sharing views as the backbone. To remedy the potential instability of the ViT, we incorporate a convolutional stem to split each augmented sample into a sequence of patches, which uses multiple stacked small convolutions instead of a big convolution in the patch projection layer. By learning the feature representations for the sequences of patches via the backbone, an instance projector and a cluster projector are further utilized to perform the instance-level contrastive learning and the global clustering structure learning, respectively. Experiments on eight image datasets demonstrate the stability (during the training-from-scratch) and the superiority (in clustering performance) of our VTCC approach over the state-of-the-art.
Self-supervised methods have significantly closed the gap with end-to-end supervised learning for image classification. In the case of human action videos, however, where both appearance and motion are significant factors of variation, this gap remains significant. One of the key reasons for this is that sampling pairs of similar video clips, a required step for many self-supervised contrastive learning methods, is currently done conservatively to avoid false positives. A typical assumption is that similar clips only occur temporally close within a single video, leading to insufficient examples of motion similarity. To mitigate this, we propose SLIC, a clustering-based self-supervised contrastive learning method for human action videos. Our key contribution is that we improve upon the traditional intra-video positive sampling by using iterative clustering to group similar video instances. This enables our method to leverage pseudo-labels from the cluster assignments to sample harder positives and negatives. SLIC outperforms state-of-the-art video retrieval baselines by +15.4% on top-1 recall on UCF101 and by +5.7% when directly transferred to HMDB51. With end-to-end finetuning for action classification, SLIC achieves 83.2% top-1 accuracy (+0.8%) on UCF101 and 54.5% on HMDB51 (+1.6%). SLIC is also competitive with the state-of-the-art in action classification after self-supervised pretraining on Kinetics400.
Establishing voxelwise semantic correspondence across distinct imaging modalities is a foundational yet formidable computer vision task. Current multi-modality registration techniques maximize hand-crafted inter-domain similarity functions, are limited in modeling nonlinear intensity-relationships and deformations, and may require significant re-engineering or underperform on new tasks, datasets, and domain pairs. This work presents ContraReg, an unsupervised contrastive representation learning approach to multi-modality deformable registration. By projecting learned multi-scale local patch features onto a jointly learned inter-domain embedding space, ContraReg obtains representations useful for non-rigid multi-modality alignment. Experimentally, ContraReg achieves accurate and robust results with smooth and invertible deformations across a series of baselines and ablations on a neonatal T1-T2 brain MRI registration task with all methods validated over a wide range of deformation regularization strengths.
Knowledge embeddings (KE) represent a knowledge graph (KG) by embedding entities and relations into continuous vector spaces. Existing methods are mainly structure-based or description-based. Structure-based methods learn representations that preserve the inherent structure of KGs. They cannot well represent abundant long-tail entities in real-world KGs with limited structural information. Description-based methods leverage textual information and language models. Prior approaches in this direction barely outperform structure-based ones, and suffer from problems like expensive negative sampling and restrictive description demand. In this paper, we propose LMKE, which adopts Language Models to derive Knowledge Embeddings, aiming at both enriching representations of long-tail entities and solving problems of prior description-based methods. We formulate description-based KE learning with a contrastive learning framework to improve efficiency in training and evaluation. Experimental results show that LMKE achieves state-of-the-art performance on KE benchmarks of link prediction and triple classification, especially for long-tail entities.
Expressive speech synthesis, like audiobook synthesis, is still challenging for style representation learning and prediction. Deriving from reference audio or predicting style tags from text requires a huge amount of labeled data, which is costly to acquire and difficult to define and annotate accurately. In this paper, we propose a novel framework for learning style representation from abundant plain text in a self-supervised manner. It leverages an emotion lexicon and uses contrastive learning and deep clustering. We further integrate the style representation as a conditioned embedding in a multi-style Transformer TTS. Comparing with multi-style TTS by predicting style tags trained on the same dataset but with human annotations, our method achieves improved results according to subjective evaluations on both in-domain and out-of-domain test sets in audiobook speech. Moreover, with implicit context-aware style representation, the emotion transition of synthesized audio in a long paragraph appears more natural. The audio samples are available on the demo web.
In this paper, we leverage the human perceiving process, that involves vision and language interaction, to generate a coherent paragraph description of untrimmed videos. We propose vision-language (VL) features consisting of two modalities, i.e., (i) vision modality to capture global visual content of the entire scene and (ii) language modality to extract scene elements description of both human and non-human objects (e.g. animals, vehicles, etc), visual and non-visual elements (e.g. relations, activities, etc). Furthermore, we propose to train our proposed VLCap under a contrastive learning VL loss. The experiments and ablation studies on ActivityNet Captions and YouCookII datasets show that our VLCap outperforms existing SOTA methods on both accuracy and diversity metrics.
Can we inject the pocket-ligand interaction knowledge into the pre-trained model and jointly learn their chemical space? Pretraining molecules and proteins has attracted considerable attention in recent years, while most of these approaches focus on learning one of the chemical spaces and lack the injection of biological knowledge. We propose a co-supervised pretraining (CoSP) framework to simultaneously learn 3D pocket and ligand representations. We use a gated geometric message passing layer to model both 3D pockets and ligands, where each node's chemical features, geometric position and orientation are considered. To learn biological meaningful embeddings, we inject the pocket-ligand interaction knowledge into the pretraining model via contrastive loss. Considering the specificity of molecules, we further propose a chemical similarity-enhanced negative sampling strategy to improve the contrastive learning performance. Through extensive experiments, we conclude that CoSP can achieve competitive results in pocket matching, molecule property predictions, and virtual screening.
Image and Point Clouds provide different information for robots. Finding the correspondences between data from different sensors is crucial for various tasks such as localization, mapping, and navigation. Learning-based descriptors have been developed for single sensors; there is little work on cross-modal features. This work treats learning cross-modal features as a dense contrastive learning problem. We propose a Tuple-Circle loss function for cross-modality feature learning. Furthermore, to learn good features and not lose generality, we developed a variant of widely used PointNet++ architecture for point cloud and U-Net CNN architecture for images. Moreover, we conduct experiments on a real-world dataset to show the effectiveness of our loss function and network structure. We show that our models indeed learn information from both images as well as LiDAR by visualizing the features.
Graph instance contrastive learning has been proved as an effective task for Graph Neural Network (GNN) pre-training. However, one key issue may seriously impede the representative power in existing works: Positive instances created by current methods often miss crucial information of graphs or even yield illegal instances (such as non-chemically-aware graphs in molecular generation). To remedy this issue, we propose to select positive graph instances directly from existing graphs in the training set, which ultimately maintains the legality and similarity to the target graphs. Our selection is based on certain domain-specific pair-wise similarity measurements as well as sampling from a hierarchical graph encoding similarity relations among graphs. Besides, we develop an adaptive node-level pre-training method to dynamically mask nodes to distribute them evenly in the graph. We conduct extensive experiments on $13$ graph classification and node classification benchmark datasets from various domains. The results demonstrate that the GNN models pre-trained by our strategies can outperform those trained-from-scratch models as well as the variants obtained by existing methods.
New astronomical tasks are often related to earlier tasks for which labels have already been collected. We adapt the contrastive framework BYOL to leverage those labels as a pretraining task while also enforcing augmentation invariance. For large-scale pretraining, we introduce GZ-Evo v0.1, a set of 96.5M volunteer responses for 552k galaxy images plus a further 1.34M comparable unlabelled galaxies. Most of the 206 GZ-Evo answers are unknown for any given galaxy, and so our pretraining task uses a Dirichlet loss that naturally handles unknown answers. GZ-Evo pretraining, with or without hybrid learning, improves on direct training even with plentiful downstream labels (+4% accuracy with 44k labels). Our hybrid pretraining/contrastive method further improves downstream accuracy vs. pretraining or contrastive learning, especially in the low-label transfer regime (+6% accuracy with 750 labels).
It has been verified that only a small fraction of the neoantigens presented by MHC class I molecules on the cell surface can elicit T cells. The limitation can be attributed to the binding specificity of T cell receptor (TCR) to peptide-MHC complex (pMHC). Computational prediction of T cell binding to neoantigens is an challenging and unresolved task. In this paper, we propose an attentive-mask contrastive learning model, ATMTCR, for inferring TCR-antigen binding specificity. For each input TCR sequence, we used Transformer encoder to transform it to latent representation, and then masked a proportion of residues guided by attention weights to generate its contrastive view. Pretraining on large-scale TCR CDR3 sequences, we verified that contrastive learning significantly improved the prediction performance of TCR binding to peptide-MHC complex (pMHC). Beyond the detection of important amino acids and their locations in the TCR sequence, our model can also extracted high-order semantic information underlying the TCR-antigen binding specificity. Comparison experiments were conducted on two independent datasets, our method achieved better performance than other existing algorithms. Moreover, we effectively identified important amino acids and their positional preferences through attention weights, which indicated the interpretability of our proposed model.
Animal pose estimation is challenging for existing image-based methods because of limited training data and large intra- and inter-species variances. Motivated by the progress of visual-language research, we propose that pre-trained language models (e.g., CLIP) can facilitate animal pose estimation by providing rich prior knowledge for describing animal keypoints in text. However, we found that building effective connections between pre-trained language models and visual animal keypoints is non-trivial since the gap between text-based descriptions and keypoint-based visual features about animal pose can be significant. To address this issue, we introduce a novel prompt-based Contrastive learning scheme for connecting Language and AniMal Pose (CLAMP) effectively. The CLAMP attempts to bridge the gap by adapting the text prompts to the animal keypoints during network training. The adaptation is decomposed into spatial-aware and feature-aware processes, and two novel contrastive losses are devised correspondingly. In practice, the CLAMP enables the first cross-modal animal pose estimation paradigm. Experimental results show that our method achieves state-of-the-art performance under the supervised, few-shot, and zero-shot settings, outperforming image-based methods by a large margin.
We present an approach that incorporates expert knowledge for time-series representation learning. Our method employs expert features to replace the commonly used data transformations in previous contrastive learning approaches. We do this since time-series data frequently stems from the industrial or medical field where expert features are often available from domain experts, while transformations are generally elusive for time-series data. We start by proposing two properties that useful time-series representations should fulfill and show that current representation learning approaches do not ensure these properties. We therefore devise ExpCLR, a novel contrastive learning approach built on an objective that utilizes expert features to encourage both properties for the learned representation. Finally, we demonstrate on three real-world time-series datasets that ExpCLR surpasses several state-of-the-art methods for both unsupervised and semi-supervised representation learning.
Recent advances in the industrial inspection of textured surfaces-in the form of visual inspection-have made such inspections possible for efficient, flexible manufacturing systems. We propose an unsupervised feature memory rearrangement network (FMR-Net) to accurately detect various textural defects simultaneously. Consistent with mainstream methods, we adopt the idea of background reconstruction; however, we innovatively utilize artificial synthetic defects to enable the model to recognize anomalies, while traditional wisdom relies only on defect-free samples. First, we employ an encoding module to obtain multiscale features of the textured surface. Subsequently, a contrastive-learning-based memory feature module (CMFM) is proposed to obtain discriminative representations and construct a normal feature memory bank in the latent space, which can be employed as a substitute for defects and fast anomaly scores at the patch level. Next, a novel global feature rearrangement module (GFRM) is proposed to further suppress the reconstruction of residual defects. Finally, a decoding module utilizes the restored features to reconstruct the normal texture background. In addition, to improve inspection performance, a two-phase training strategy is utilized for accurate defect restoration refinement, and we exploit a multimodal inspection method to achieve noise-robust defect localization. We verify our method through extensive experiments and test its practical deployment in collaborative edge--cloud intelligent manufacturing scenarios by means of a multilevel detection method, demonstrating that FMR-Net exhibits state-of-the-art inspection accuracy and shows great potential for use in edge-computing-enabled smart industries.
In this work, we introduce panoramic panoptic segmentation, as the most holistic scene understanding, both in terms of Field of View (FoV) and image-level understanding for standard camera-based input. A complete surrounding understanding provides a maximum of information to a mobile agent. This is essential information for any intelligent vehicle to make informed decisions in a safety-critical dynamic environment such as real-world traffic. In order to overcome the lack of annotated panoramic images, we propose a framework which allows model training on standard pinhole images and transfers the learned features to the panoramic domain in a cost-minimizing way. The domain shift from pinhole to panoramic images is non-trivial as large objects and surfaces are heavily distorted close to the image border regions and look different across the two domains. Using our proposed method with dense contrastive learning, we manage to achieve significant improvements over a non-adapted approach. Depending on the efficient panoptic segmentation architecture, we can improve 3.5-6.5% measured in Panoptic Quality (PQ) over non-adapted models on our established Wild Panoramic Panoptic Segmentation (WildPPS) dataset. Furthermore, our efficient framework does not need access to the images of the target domain, making it a feasible domain generalization approach suitable for a limited hardware setting. As additional contributions, we publish WildPPS: The first panoramic panoptic image dataset to foster progress in surrounding perception and explore a novel training procedure combining supervised and contrastive training.
We study offline meta-reinforcement learning, a practical reinforcement learning paradigm that learns from offline data to adapt to new tasks. The distribution of offline data is determined jointly by the behavior policy and the task. Existing offline meta-reinforcement learning algorithms cannot distinguish these factors, making task representations unstable to the change of behavior policies. To address this problem, we propose a contrastive learning framework for task representations that are robust to the distribution mismatch of behavior policies in training and test. We design a bi-level encoder structure, use mutual information maximization to formalize task representation learning, derive a contrastive learning objective, and introduce several approaches to approximate the true distribution of negative pairs. Experiments on a variety of offline meta-reinforcement learning benchmarks demonstrate the advantages of our method over prior methods, especially on the generalization to out-of-distribution behavior policies. The code is available at https://github.com/PKU-AI-Edge/CORRO.
Learning-based multi-view stereo (MVS) methods have made impressive progress and surpassed traditional methods in recent years. However, their accuracy and completeness are still struggling. In this paper, we propose a new method to enhance the performance of existing networks inspired by contrastive learning and feature matching. First, we propose a Contrast Matching Loss (CML), which treats the correct matching points in depth-dimension as positive sample and other points as negative samples, and computes the contrastive loss based on the similarity of features. We further propose a Weighted Focal Loss (WFL) for better classification capability, which weakens the contribution of low-confidence pixels in unimportant areas to the loss according to predicted confidence. Extensive experiments performed on DTU, Tanks and Temples and BlendedMVS datasets show our method achieves state-of-the-art performance and significant improvement over baseline network.
Video highlight detection is a crucial yet challenging problem that aims to identify the interesting moments in untrimmed videos. The key to this task lies in effective video representations that jointly pursue two goals, \textit{i.e.}, cross-modal representation learning and fine-grained feature discrimination. In this paper, these two challenges are tackled by not only enriching intra-modality and cross-modality relations for representation modeling but also shaping the features in a discriminative manner. Our proposed method mainly leverages the intra-modality encoding and cross-modality co-occurrence encoding for fully representation modeling. Specifically, intra-modality encoding augments the modality-wise features and dampens irrelevant modality via within-modality relation learning in both audio and visual signals. Meanwhile, cross-modality co-occurrence encoding focuses on the co-occurrence inter-modality relations and selectively captures effective information among multi-modality. The multi-modal representation is further enhanced by the global information abstracted from the local context. In addition, we enlarge the discriminative power of feature embedding with a hard-pairs guided contrastive learning (HPCL) scheme. A hard-pairs sampling strategy is further employed to mine the hard samples for improving feature discrimination in HPCL. Extensive experiments conducted on two benchmarks demonstrate the effectiveness and superiority of our proposed methods compared to other state-of-the-art methods.
Unsupervised image retrieval aims to learn an efficient retrieval system without expensive data annotations, but most existing methods rely heavily on handcrafted feature descriptors or pre-trained feature extractors. To minimize human supervision, recent advance proposes deep fully unsupervised image retrieval aiming at training a deep model from scratch to jointly optimize visual features and quantization codes. However, existing approach mainly focuses on instance contrastive learning without considering underlying semantic structure information, resulting in sub-optimal performance. In this work, we propose a novel self-supervised consistent quantization approach to deep fully unsupervised image retrieval, which consists of part consistent quantization and global consistent quantization. In part consistent quantization, we devise part neighbor semantic consistency learning with codeword diversity regularization. This allows to discover underlying neighbor structure information of sub-quantized representations as self-supervision. In global consistent quantization, we employ contrastive learning for both embedding and quantized representations and fuses these representations for consistent contrastive regularization between instances. This can make up for the loss of useful representation information during quantization and regularize consistency between instances. With a unified learning objective of part and global consistent quantization, our approach exploits richer self-supervision cues to facilitate model learning. Extensive experiments on three benchmark datasets show the superiority of our approach over the state-of-the-art methods.