Audio-visual segmentation (AVS) is a challenging task that involves accurately segmenting sounding objects based on audio-visual cues. The effectiveness of audio-visual learning critically depends on achieving accurate cross-modal alignment between sound and visual objects. Successful audio-visual learning requires two essential components: 1) a challenging dataset with high-quality pixel-level multi-class annotated images associated with audio files, and 2) a model that can establish strong links between audio information and its corresponding visual object. However, these requirements are only partially addressed by current methods, with training sets containing biased audio-visual data, and models that generalise poorly beyond this biased training set. In this work, we propose a new cost-effective strategy to build challenging and relatively unbiased high-quality audio-visual segmentation benchmarks. We also propose a new informative sample mining method for audio-visual supervised contrastive learning to leverage discriminative contrastive samples to enforce cross-modal understanding. We show empirical results that demonstrate the effectiveness of our benchmark. Furthermore, experiments conducted on existing AVS datasets and on our new benchmark show that our method achieves state-of-the-art (SOTA) segmentation accuracy.
The long-standing goal of dense retrievers in abtractive open-domain question answering (ODQA) tasks is to learn to capture evidence passages among relevant passages for any given query, such that the reader produce factually correct outputs from evidence passages. One of the key challenge is the insufficient amount of training data with the supervision of the answerability of the passages. Recent studies rely on iterative pipelines to annotate answerability using signals from the reader, but their high computational costs hamper practical applications. In this paper, we instead focus on a data-centric approach and propose Evidentiality-Aware Dense Passage Retrieval (EADPR), which leverages synthetic distractor samples to learn to discriminate evidence passages from distractors. We conduct extensive experiments to validate the effectiveness of our proposed method on multiple abstractive ODQA tasks.
Medical data often exhibits long-tail distributions with heavy class imbalance, which naturally leads to difficulty in classifying the minority classes (i.e., boundary regions or rare objects). Recent work has significantly improved semi-supervised medical image segmentation in long-tailed scenarios by equipping them with unsupervised contrastive criteria. However, it remains unclear how well they will perform in the labeled portion of data where class distribution is also highly imbalanced. In this work, we present ACTION++, an improved contrastive learning framework with adaptive anatomical contrast for semi-supervised medical segmentation. Specifically, we propose an adaptive supervised contrastive loss, where we first compute the optimal locations of class centers uniformly distributed on the embedding space (i.e., off-line), and then perform online contrastive matching training by encouraging different class features to adaptively match these distinct and uniformly distributed class centers. Moreover, we argue that blindly adopting a constant temperature $\tau$ in the contrastive loss on long-tailed medical data is not optimal, and propose to use a dynamic $\tau$ via a simple cosine schedule to yield better separation between majority and minority classes. Empirically, we evaluate ACTION++ on ACDC and LA benchmarks and show that it achieves state-of-the-art across two semi-supervised settings. Theoretically, we analyze the performance of adaptive anatomical contrast and confirm its superiority in label efficiency.
In computer vision, contrastive learning is the most advanced unsupervised learning framework. Yet most previous methods simply apply fixed composition of data augmentations to improve data efficiency, which ignores the changes in their optimal settings over training. Thus, the pre-determined parameters of augmentation operations cannot always fit well with an evolving network during the whole training period, which degrades the quality of the learned representations. In this work, we propose AdDA, which implements a closed-loop feedback structure to a generic contrastive learning network. AdDA works by allowing the network to adaptively adjust the augmentation compositions according to the real-time feedback. This online adjustment helps maintain the dynamic optimal composition and enables the network to acquire more generalizable representations with minimal computational overhead. AdDA achieves competitive results under the common linear protocol on ImageNet-100 classification (+1.11% on MoCo v2).
Misinformation has become a pressing issue. Fake media, in both visual and textual forms, is widespread on the web. While various deepfake detection and text fake news detection methods have been proposed, they are only designed for single-modality forgery based on binary classification, let alone analyzing and reasoning subtle forgery traces across different modalities. In this paper, we highlight a new research problem for multi-modal fake media, namely Detecting and Grounding Multi-Modal Media Manipulation (DGM^4). DGM^4 aims to not only detect the authenticity of multi-modal media, but also ground the manipulated content (i.e., image bounding boxes and text tokens), which requires deeper reasoning of multi-modal media manipulation. To support a large-scale investigation, we construct the first DGM^4 dataset, where image-text pairs are manipulated by various approaches, with rich annotation of diverse manipulations. Moreover, we propose a novel HierArchical Multi-modal Manipulation rEasoning tRansformer (HAMMER) to fully capture the fine-grained interaction between different modalities. HAMMER performs 1) manipulation-aware contrastive learning between two uni-modal encoders as shallow manipulation reasoning, and 2) modality-aware cross-attention by multi-modal aggregator as deep manipulation reasoning. Dedicated manipulation detection and grounding heads are integrated from shallow to deep levels based on the interacted multi-modal information. Finally, we build an extensive benchmark and set up rigorous evaluation metrics for this new research problem. Comprehensive experiments demonstrate the superiority of our model; several valuable observations are also revealed to facilitate future research in multi-modal media manipulation.
The truth is significantly hampered by massive rumors that spread along with breaking news or popular topics. Since there is sufficient corpus gathered from the same domain for model training, existing rumor detection algorithms show promising performance on yesterday's news. However, due to a lack of substantial training data and prior expert knowledge, they are poor at spotting rumors concerning unforeseen events, especially those propagated in different languages (i.e., low-resource regimes). In this paper, we propose a unified contrastive transfer framework to detect rumors by adapting the features learned from well-resourced rumor data to that of the low-resourced with only few-shot annotations. More specifically, we first represent rumor circulated on social media as an undirected topology for enhancing the interaction of user opinions, and then train a Multi-scale Graph Convolutional Network via a unified contrastive paradigm to mine effective clues simultaneously from post semantics and propagation structure. Our model explicitly breaks the barriers of the domain and/or language issues, via language alignment and a novel domain-adaptive contrastive learning mechanism. To well-generalize the representation learning using a small set of annotated target events, we reveal that rumor-indicative signal is closely correlated with the uniformity of the distribution of these events. We design a target-wise contrastive training mechanism with three event-level data augmentation strategies, capable of unifying the representations by distinguishing target events. Extensive experiments conducted on four low-resource datasets collected from real-world microblog platforms demonstrate that our framework achieves much better performance than state-of-the-art methods and exhibits a superior capacity for detecting rumors at early stages.
Federated learning-assisted edge intelligence enables privacy protection in modern intelligent services. However, not independent and identically distributed (non-IID) distribution among edge clients can impair the local model performance. The existing single prototype-based strategy represents a class by using the mean of the feature space. However, feature spaces are usually not clustered, and a single prototype may not represent a class well. Motivated by this, this paper proposes a multi-prototype federated contrastive learning approach (MP-FedCL) which demonstrates the effectiveness of using a multi-prototype strategy over a single-prototype under non-IID settings, including both label and feature skewness. Specifically, a multi-prototype computation strategy based on \textit{k-means} is first proposed to capture different embedding representations for each class space, using multiple prototypes ($k$ centroids) to represent a class in the embedding space. In each global round, the computed multiple prototypes and their respective model parameters are sent to the edge server for aggregation into a global prototype pool, which is then sent back to all clients to guide their local training. Finally, local training for each client minimizes their own supervised learning tasks and learns from shared prototypes in the global prototype pool through supervised contrastive learning, which encourages them to learn knowledge related to their own class from others and reduces the absorption of unrelated knowledge in each global iteration. Experimental results on MNIST, Digit-5, Office-10, and DomainNet show that our method outperforms multiple baselines, with an average test accuracy improvement of about 4.6\% and 10.4\% under feature and label non-IID distributions, respectively.
Vision-Language (V-L) models trained with contrastive learning to align the visual and language modalities have been shown to be strong few-shot learners. Soft prompt learning is the method of choice for few-shot downstream adaptation aiming to bridge the modality gap caused by the distribution shift induced by the new domain. While parameter-efficient, prompt learning still requires access to the model weights and can be computationally infeasible for large models with billions of parameters. To address these shortcomings, in this work, we describe a black-box method for V-L few-shot adaptation that (a) operates on pre-computed image and text features and hence works without access to the model's weights, (b) it is orders of magnitude faster at training time, (c) it is amenable to both supervised and unsupervised training, and (d) it can be even used to align image and text features computed from uni-modal models. To achieve this, we propose Linear Feature Alignment (LFA), a simple linear approach for V-L re-alignment in the target domain. LFA is initialized from a closed-form solution to a least-squares problem and then it is iteratively updated by minimizing a re-ranking loss. Despite its simplicity, our approach can even surpass soft-prompt learning methods as shown by extensive experiments on 11 image and 2 video datasets.
Modern data augmentation using a mixture-based technique can regularize the models from overfitting to the training data in various computer vision applications, but a proper data augmentation technique tailored for the part-based Visible-Infrared person Re-IDentification (VI-ReID) models remains unexplored. In this paper, we present a novel data augmentation technique, dubbed PartMix, that synthesizes the augmented samples by mixing the part descriptors across the modalities to improve the performance of part-based VI-ReID models. Especially, we synthesize the positive and negative samples within the same and across different identities and regularize the backbone model through contrastive learning. In addition, we also present an entropy-based mining strategy to weaken the adverse impact of unreliable positive and negative samples. When incorporated into existing part-based VI-ReID model, PartMix consistently boosts the performance. We conduct experiments to demonstrate the effectiveness of our PartMix over the existing VI-ReID methods and provide ablation studies.
Abnormal event detection, which refers to mining unusual interactions among involved entities, plays an important role in many real applications. Previous works mostly over-simplify this task as detecting abnormal pair-wise interactions. However, real-world events may contain multi-typed attributed entities and complex interactions among them, which forms an Attributed Heterogeneous Information Network (AHIN). With the boom of social networks, abnormal event detection in AHIN has become an important, but seldom explored task. In this paper, we firstly study the unsupervised abnormal event detection problem in AHIN. The events are considered as star-schema instances of AHIN and are further modeled by hypergraphs. A novel hypergraph contrastive learning method, named AEHCL, is proposed to fully capture abnormal event patterns. AEHCL designs the intra-event and inter-event contrastive modules to exploit self-supervised AHIN information. The intra-event contrastive module captures the pair-wise and multivariate interaction anomalies within an event, and the inter-event module captures the contextual anomalies among events. These two modules collaboratively boost the performance of each other and improve the detection results. During the testing phase, a contrastive learning-based abnormal event score function is further proposed to measure the abnormality degree of events. Extensive experiments on three datasets in different scenarios demonstrate the effectiveness of AEHCL, and the results improve state-of-the-art baselines up to 12.0% in Average Precision (AP) and 4.6% in Area Under Curve (AUC) respectively.
The deep neural networks (DNNs) have freed the synthetic aperture radar automatic target recognition (SAR ATR) from expertise-based feature designing and demonstrated superiority over conventional solutions. There has been shown the unique deficiency of ground vehicle benchmarks in shapes of strong background correlation results in DNNs overfitting the clutter and being non-robust to unfamiliar surroundings. However, the gap between fixed background model training and varying background application remains underexplored. Inspired by contrastive learning, this letter proposes a solution called Contrastive Feature Alignment (CFA) aiming to learn invariant representation for robust recognition. The proposed method contributes a mixed clutter variants generation strategy and a new inference branch equipped with channel-weighted mean square error (CWMSE) loss for invariant representation learning. In specific, the generation strategy is delicately designed to better attract clutter-sensitive deviation in feature space. The CWMSE loss is further devised to better contrast this deviation and align the deep features activated by the original images and corresponding clutter variants. The proposed CFA combines both classification and CWMSE losses to train the model jointly, which allows for the progressive learning of invariant target representation. Extensive evaluations on the MSTAR dataset and six DNN models prove the effectiveness of our proposal. The results demonstrated that the CFA-trained models are capable of recognizing targets among unfamiliar surroundings that are not included in the dataset, and are robust to varying signal-to-clutter ratios.
In this work, we present Multi-Level Contrastive Learning for Dense Prediction Task (MCL), an efficient self-supervised method for learning region-level feature representation for dense prediction tasks. Our method is motivated by the three key factors in detection: localization, scale consistency and recognition. To explicitly encode absolute position and scale information, we propose a novel pretext task that assembles multi-scale images in a montage manner to mimic multi-object scenarios. Unlike the existing image-level self-supervised methods, our method constructs a multi-level contrastive loss that considers each sub-region of the montage image as a singleton. Our method enables the neural network to learn regional semantic representations for translation and scale consistency while reducing pre-training epochs to the same as supervised pre-training. Extensive experiments demonstrate that MCL consistently outperforms the recent state-of-the-art methods on various datasets with significant margins. In particular, MCL obtains 42.5 AP$^\mathrm{bb}$ and 38.3 AP$^\mathrm{mk}$ on COCO with the 1x schedule fintuning, when using Mask R-CNN with R50-FPN backbone pre-trained with 100 epochs. In comparison to MoCo, our method surpasses their performance by 4.0 AP$^\mathrm{bb}$ and 3.1 AP$^\mathrm{mk}$. Furthermore, we explore the alignment between pretext task and downstream tasks. We extend our pretext task to supervised pre-training, which achieves a similar performance to self-supervised learning. This result demonstrates the importance of the alignment between pretext task and downstream tasks, indicating the potential for wider applicability of our method beyond self-supervised settings.
Document-level relation extraction (DocRE) predicts relations for entity pairs that rely on long-range context-dependent reasoning in a document. As a typical multi-label classification problem, DocRE faces the challenge of effectively distinguishing a small set of positive relations from the majority of negative ones. This challenge becomes even more difficult to overcome when there exists a significant number of annotation errors in the dataset. In this work, we aim to achieve better integration of both the discriminability and robustness for the DocRE problem. Specifically, we first design an effective loss function to endow high discriminability to both probabilistic outputs and internal representations. We innovatively customize entropy minimization and supervised contrastive learning for the challenging multi-label and long-tailed learning problems. To ameliorate the impact of label errors, we equipped our method with a novel negative label sampling strategy to strengthen the model robustness. In addition, we introduce two new data regimes to mimic more realistic scenarios with annotation errors and evaluate our sampling strategy. Experimental results verify the effectiveness of each component and show that our method achieves new state-of-the-art results on the DocRED dataset, its recently cleaned version, Re-DocRED, and the proposed data regimes.
Contrastive learning is an efficient approach to self-supervised representation learning. Although recent studies have made progress in the theoretical understanding of contrastive learning, the investigation of how to characterize the clusters of the learned representations is still limited. In this paper, we aim to elucidate the characterization from theoretical perspectives. To this end, we consider a kernel-based contrastive learning framework termed Kernel Contrastive Learning (KCL), where kernel functions play an important role when applying our theoretical results to other frameworks. We introduce a formulation of the similarity structure of learned representations by utilizing a statistical dependency viewpoint. We investigate the theoretical properties of the kernel-based contrastive loss via this formulation. We first prove that the formulation characterizes the structure of representations learned with the kernel-based contrastive learning framework. We show a new upper bound of the classification error of a downstream task, which explains that our theory is consistent with the empirical success of contrastive learning. We also establish a generalization error bound of KCL. Finally, we show a guarantee for the generalization ability of KCL to the downstream classification task via a surrogate bound.
Current state-of-the-art approaches for few-shot action recognition achieve promising performance by conducting frame-level matching on learned visual features. However, they generally suffer from two limitations: i) the matching procedure between local frames tends to be inaccurate due to the lack of guidance to force long-range temporal perception; ii) explicit motion learning is usually ignored, leading to partial information loss. To address these issues, we develop a Motion-augmented Long-short Contrastive Learning (MoLo) method that contains two crucial components, including a long-short contrastive objective and a motion autodecoder. Specifically, the long-short contrastive objective is to endow local frame features with long-form temporal awareness by maximizing their agreement with the global token of videos belonging to the same class. The motion autodecoder is a lightweight architecture to reconstruct pixel motions from the differential features, which explicitly embeds the network with motion dynamics. By this means, MoLo can simultaneously learn long-range temporal context and motion cues for comprehensive few-shot matching. To demonstrate the effectiveness, we evaluate MoLo on five standard benchmarks, and the results show that MoLo favorably outperforms recent advanced methods. The source code is available at https://github.com/alibaba-mmai-research/MoLo.
Learning view-invariant representation is a key to improving feature discrimination power for skeleton-based action recognition. Existing approaches cannot effectively remove the impact of viewpoint due to the implicit view-dependent representations. In this work, we propose a self-supervised framework called Focalized Contrastive View-invariant Learning (FoCoViL), which significantly suppresses the view-specific information on the representation space where the viewpoints are coarsely aligned. By maximizing mutual information with an effective contrastive loss between multi-view sample pairs, FoCoViL associates actions with common view-invariant properties and simultaneously separates the dissimilar ones. We further propose an adaptive focalization method based on pairwise similarity to enhance contrastive learning for a clearer cluster boundary in the learned space. Different from many existing self-supervised representation learning work that rely heavily on supervised classifiers, FoCoViL performs well on both unsupervised and supervised classifiers with superior recognition performance. Extensive experiments also show that the proposed contrastive-based focalization generates a more discriminative latent representation.
The goal of open-vocabulary detection is to identify novel objects based on arbitrary textual descriptions. In this paper, we address open-vocabulary 3D point-cloud detection by a dividing-and-conquering strategy, which involves: 1) developing a point-cloud detector that can learn a general representation for localizing various objects, and 2) connecting textual and point-cloud representations to enable the detector to classify novel object categories based on text prompting. Specifically, we resort to rich image pre-trained models, by which the point-cloud detector learns localizing objects under the supervision of predicted 2D bounding boxes from 2D pre-trained detectors. Moreover, we propose a novel de-biased triplet cross-modal contrastive learning to connect the modalities of image, point-cloud and text, thereby enabling the point-cloud detector to benefit from vision-language pre-trained models,i.e.,CLIP. The novel use of image and vision-language pre-trained models for point-cloud detectors allows for open-vocabulary 3D object detection without the need for 3D annotations. Experiments demonstrate that the proposed method improves at least 3.03 points and 7.47 points over a wide range of baselines on the ScanNet and SUN RGB-D datasets, respectively. Furthermore, we provide a comprehensive analysis to explain why our approach works.
We propose a visual-linguistic representation learning approach within a self-supervised learning framework by introducing a new operation, loss, and data augmentation strategy. First, we generate diverse features for the image-text matching (ITM) task via soft-masking the regions in an image, which are most relevant to a certain word in the corresponding caption, instead of completely removing them. Since our framework relies only on image-caption pairs with no fine-grained annotations, we identify the relevant regions to each word by computing the word-conditional visual attention using multi-modal encoder. Second, we encourage the model to focus more on hard but diverse examples by proposing a focal loss for the image-text contrastive learning (ITC) objective, which alleviates the inherent limitations of overfitting and bias issues. Last, we perform multi-modal data augmentations for self-supervised learning via mining various examples by masking texts and rendering distortions on images. We show that the combination of these three innovations is effective for learning a pretrained model, leading to outstanding performance on multiple vision-language downstream tasks.
Transformations based on domain expertise (expert transformations), such as random-resized-crop and color-jitter, have proven critical to the success of contrastive learning techniques such as SimCLR. Recently, several attempts have been made to replace such domain-specific, human-designed transformations with generated views that are learned. However for imagery data, so far none of these view-generation methods has been able to outperform expert transformations. In this work, we tackle a different question: instead of replacing expert transformations with generated views, can we constructively assimilate generated views with expert transformations? We answer this question in the affirmative and propose a view generation method and a simple, effective assimilation method that together improve the state-of-the-art by up to ~3.6% on three different datasets. Importantly, we conduct a detailed empirical study that systematically analyzes a range of view generation and assimilation methods and provides a holistic picture of the efficacy of learned views in contrastive representation learning.
Learning similarity between scene graphs and images aims to estimate a similarity score given a scene graph and an image. There is currently no research dedicated to this task, although it is critical for scene graph generation and downstream applications. Scene graph generation is conventionally evaluated by Recall$@K$ and mean Recall$@K$, which measure the ratio of predicted triplets that appear in the human-labeled triplet set. However, such triplet-oriented metrics fail to demonstrate the overall semantic difference between a scene graph and an image and are sensitive to annotation bias and noise. Using generated scene graphs in the downstream applications is therefore limited. To address this issue, for the first time, we propose a Scene graPh-imAge coNtrastive learning framework, SPAN, that can measure the similarity between scene graphs and images. Our novel framework consists of a graph Transformer and an image Transformer to align scene graphs and their corresponding images in the shared latent space. We introduce a novel graph serialization technique that transforms a scene graph into a sequence with structural encodings. Based on our framework, we propose R-Precision measuring image retrieval accuracy as a new evaluation metric for scene graph generation. We establish new benchmarks on the Visual Genome and Open Images datasets. Extensive experiments are conducted to verify the effectiveness of SPAN, which shows great potential as a scene graph encoder.