Cross-modal retrieval is a fundamental application of multi-modal learning that has achieved remarkable success with large-scale well-paired data. However, in practice, it is costly to collect large-scale well-paired data. To alleviate the dependence on the amount of paired data, in this paper, we study a practical learning paradigm: semi-paired cross-modal learning (SPL), which utilizes both a small amount of paired data and a large amount of unpaired data to enhance cross-modal learning directly and is more accessible in practice. To achieve this, we take image-text retrieval as an example and propose a novel Robust Cross-modal Semi-paired Learning method (RCSL) by addressing two challenges. To be specific, i) to overcome the under-optimization issue caused by too little paired data, we present Semi-paired Discriminative Learning (SDL) to fully learn visual-semantic associations from a small amount of image-text pairs by preserving the alignment and uniformity of modality representations. ii) To mine visual-semantic correspondences from unpaired data, RCSL first constructs pseudo-paired correlations across different modalities by nearest neighbor association. However, this may introduce noisy correspondences (NCs) due to inaccurate pseudo signals, which could degrade the model's performance. To tackle NCs, we devise Robust Cross-correlation Mining (RCM) based on the risk minimization criterion to robustly and explicitly learn visual-semantic associations from pseudo-paired data, thus boosting cross-modal learning. Finally, we conduct extensive experiments on four datasets, i.e., three widely used benchmark datasets of Flickr30K, MS-COCO, CC152K, and a newly constructed real-world dataset Drone-SP, to demonstrate the effectiveness of RCSL under semi-paired and noisy settings.

Due to the complexity of data collection in the real world, Multi-view Representation Learning (MvRL) always encounters the incomplete information challenge, typically manifested as the Sample-missing Problem (SP) and the View-unaligned Problem (VP). Although several methods have been proposed, they fail to find a good trade-off among sample restoration, view alignment, and data diversity preservation. To address this issue, we take and mathematically formulate two sociological concepts for MvRL, i.e., community commonality and community versatility, where the former refers to the identical custom shared within the same community, and the latter refers to the similar but non-identical custom within communities of the same minority. One could find that the community commonality can enhance the compactness of view-specific clusters, and the community versatility can preserve the view diversity. Moreover, combining both of them could facilitate achieving robust MvRL with incomplete information. With the formulations, we propose a novel method dubbed Community-Aware Multi-viEw RepresentAtion learning with incomplete information (CAMERA). In brief, CAMERA employs a novel dual-stream network and an elaborate objective function that theoretically and empirically embraces community commonality and versatility. Extensive experimental results on seven datasets demonstrate that CAMERA remarkably outperforms 24 competitive multi-view learning methods on clustering, classification, and human action recognition tasks.

The success of existing graph matching methods heavily relies on high-quality training data with complete and precise correspondences between keypoints across different graphs. However, this assumption is often violated in real-world scenarios, leading to partial correspondence and noisy correspondence challenges. In brief, partial correspondence arises from viewpoint occlusions, where certain keypoints (i.e., outliers) lack valid counterparts in the target graph, while noisy correspondence refers to both incorrectly established (i.e., false positives) and neglected (i.e., false negatives) correspondences due to annotation error. In this paper, we propose the first unified framework to address both partial and noisy correspondence challenges in graph matching. Specifically, we introduce a dual-expert cooperative framework that integrates Koopmans-Beckmann and Lawler's quadratic assignment programming formulations (KB-QAP and L-QAP) through an align-fuse-refine pipeline. In the alignment stage, the KB-QAP expert aligns keypoints and distinguishes inliers from outliers using a novel quadratic contrastive loss. In the fusion stage, the L-QAP expert employs a graph transformer on the association graph to merge the aligned graphs and incorporates a learnable outlier-rejection mechanism to handle partial correspondences. Finally, by exploiting the different noise resistances of the two experts, we identify and refine the false positive and false negative correspondences, thereby enhancing robustness against noisy correspondence. Extensive experiments on four widely-used graph matching datasets demonstrate the effectiveness of our method against 17 competitive baselines in both partial and noisy correspondence scenarios.

Multimodal learning aims to integrate diverse data sources to capture more comprehensive information about things, thus enhancing perception and understanding of the real world. However, inherent discrepancies between different modalities often lead to imbalanced optimization during multimodal learning, hindering performance improvement. To address this issue, in this paper, we present a Multimodal Information Balance (MIB) theory, grounded in Information Theory, to reveal that this imbalance arises from the imbalanced retention of complementary information during modality fusion, providing an intuitive and explainable perspective on the issue. Building on this insight, we propose a theoretical MIB criterion to adaptively balance the preservation of complementary information across individual modalities, thereby facilitating multimodal fusion. Using this criterion, we develop an Information-Balanced Multimodal Learning (IBML) framework to mine comprehensive and balanced multimodal information, achieving optimal learning. More specifically, IBML introduces Balance Information Optimization (BIO) module to maximize tractable lower bound objectives derived from the MIB criterion according to the optimization discrepancies across modalities, ensuring balanced retention of complementary information and enhancing information contributions during multimodal fusion. In addition, we present a supplementary and provable Task Complexity Modulation (TCM) module based on the MIB criterion to adjust task complexity discrepancies across input modalities, thus indirectly promoting the balanced preservation of complementary information throughout the learning process. Extensive experiments are conducted on eight multimodal datasets, spanning audio-visual recognition, image-text classification, and 2D-3D recognition, to verify the superiority and effectiveness of IBML.

With the increasing accessibility of multimodal data, cross-modal retrieval (CMR) has gained significant attention in recent years. However, most existing CMR methods are built on clean annotations and closed-set label space assumptions, which are often violated in practice. In realistic scenarios, annotations are often noisy due to machine-generated or non-expert labeling, while new categories may also emerge from heterogeneous data sources. The coexistence of label noise and open-set categories gives rise to open-set noisy labels (OSNL). Compared to closed-set noise, OSNL is more harmful because it arises from samples whose true categories lie outside the training label space. When such unknown-class samples are incorrectly assigned to known labels, the model cannot correct them through label relationships. Instead, the model is forced to learn erroneous semantic associations, embedding unknown semantics into incorrect categories. This bias gradually accumulates and disrupts the semantic structure of the shared representation space, ultimately causing existing CMR methods to struggle to maintain reliable performance. To address these challenges, this paper proposes NOise-TOlerate evidential learning (NOTO), a novel framework that robustly learns cross-modal representations under both closed-set and open-set noisy labels. Specifically, a Robust Evidential Learning (REL) module is proposed to detect clean, closed-set noisy, and open-set noisy instances by modeling the predictive distribution as Dirichlet evidence and inferring belief masses. Based on these inferred instance types, REL then assigns tailored optimization strategies to enhance semantic consistency and enlarge the discrimination margin between in-distribution data and open-set categories. An Adaptive Noise-aware Contrast (ANC) module is proposed to adaptively select reliable positive pairs according to the estimated noise states and maximize the mutual information between them to strengthen cross-modal alignment and mitigate the adverse effects of noisy supervision simultaneously. Extensive experiments and comparisons with ten state-of-the-art CMR methods on four benchmarks demonstrate that NOTO achieves superior retrieval performance and robustness against open-set noisy labels. The code is available at https://github.com/perquisite/NOTO.

Cross-modal hashing provides an efficient solution for retrieval tasks across various modalities, such as images and text. However, most existing methods are deterministic models, which overlook the reliability associated with the retrieved results. This omission renders them unreliable for determining matches between data pairs based solely on Hamming distance. To bridge the gap, in this paper, we propose a novel method called Deep Evidential Cross-modal Hashing (DECH). This method equips hashing models with the ability to quantify the reliability level of the association between a query sample and each corresponding retrieved sample, bringing a new dimension of reliability to the cross-modal retrieval process. To achieve this, our method addresses two key challenges: i) To leverage evidential theory in guiding the model to learn hash codes, we design a novel evidence acquisition module to collect evidence and place the evidence captured by hash codes on a Beta distribution to derive a binomial opinion. Unlike existing evidential learning approaches that rely on classifiers, our method collects evidence directly through hash codes. ii) To tackle the task-oriented challenge, we first introduce a method to update the derived binomial opinion, allowing it to present the uncertainty caused by conflicting evidence. Following this manner, we present a strategy to precisely evaluate the reliability level of retrieved results, culminating in performance improvement. We validate the efficacy of our DECH through extensive experimentation on four benchmark datasets. The experimental results demonstrate our superior performance compared to 12 state-of-the-art methods.

Cross-modal retrieval aims to match related samples across distinct modalities, facilitating the retrieval and discovery of heterogeneous information. Although existing methods show promising performance, most are deterministic models and are unable to capture the uncertainty inherent in the retrieval outputs, leading to potentially unreliable results. To address this issue, we propose a novel framework called FUzzy Multimodal lEarning (FUME), which is able to self-estimate epistemic uncertainty, thereby embracing trusted cross-modal retrieval. Specifically, our FUME leverages the Fuzzy Set Theory to view the outputs of the classification network as a set of membership degrees and quantify category credibility by incorporating both possibility and necessity measures. However, directly optimizing the category credibility could mislead the model by over-optimizing the necessity for unmatched categories. To overcome this challenge, we present a novel fuzzy multimodal learning strategy, which utilizes label information to guide necessity optimization in the right direction, thereby indirectly optimizing category credibility and achieving accurate decision uncertainty quantification. Furthermore, we design an uncertainty merging scheme that accounts for decision uncertainties, thus further refining uncertainty estimates and boosting the trustworthiness of retrieval results. Extensive experiments on five benchmark datasets demonstrate that FUME remarkably improves both retrieval performance and reliability, offering a prospective solution for cross-modal retrieval in high-stakes applications. Code is available at https://github.com/siyuancncd/FUME.

Composed Image Retrieval (CIR) enables editable image search by integrating a query pair—a reference image ref and a textual modification mod—to retrieve a target image tar that reflects the intended change. While existing CIR methods have shown promising performance using well-annotated triplets ⟨ref,mod,tar⟩, almost all of them implicitly assume these triplets are accurately associated with each other. In practice, however, this assumption is often violated due to the limited knowledge of annotators, inevitably leading to incorrect textual modifications and resulting in a practical yet less-touched problem: noisy triplet correspondence (NTC). To tackle this challenge, we propose a Task-oriented Modification Enhancement framework (TME) to learn robustly from noisy triplets, which comprises three key modules: Robust Fusion Query (RFQ), Pseudo Text Enhancement (PTE), and Task-Oriented Prompt (TOP). Specifically, to mitigate the adverse impact of noise, RFQ employs a sample selection strategy to divide the training triplets into clean and noisy sets, thus enhancing the reliability of the training data for robust learning. To further leverage the noisy data instead of discarding it, PTE unifies the triplet noise as an adapter mismatch problem, thereby adjusting mod to align with ref and tar in the mismatched triplet. Finally, TOP replaces ref in the clean set with a trainable prompt, which is then concatenated with mod to form a query independent of the visual reference, aiming to mitigate visually irrelevant noise. Extensive experiments on two domain-specific datasets demonstrate the robustness and superiority of TME, particularly in noisy scenarios.

Recent advancements in Mamba have shown promising results in image restoration. These methods typically flatten 2D images into multiple distinct 1D sequences along rows and columns, process each sequence independently using selective scan operation, and recombine them to form the outputs. However, such a paradigm overlooks two vital aspects: i) the local relationships and spatial continuity inherent in natural images, and ii) the discrepancies among sequences unfolded through totally different ways. To overcome the drawbacks, we explore two problems in Mamba-based restoration methods: i) how to design a scanning strategy preserving both locality and continuity while facilitating restoration, and ii) how to aggregate the distinct sequences unfolded in totally different ways. To address these problems, we propose a novel Mamba-based Image Restoration model (MaIR), which consists of Nested S-shaped Scanning strategy (NSS) and Sequence Shuffle Attention block (SSA). Specifically, NSS preserves locality and continuity of the input images through the stripe-based scanning region and the S-shaped scanning path, respectively. SSA aggregates sequences through calculating attention weights within the corresponding channels of different sequences. Thanks to NSS and SSA, MaIR surpasses 40 baselines across 14 challenging datasets, achieving state-of-the-art performance on the tasks of image super-resolution, denoising, deblurring and dehazing. The code is available at https://github.com/XLearning-SCU/2025-CVPR-MaIR.

Despite remarkable advancements in text-to-image person re-identification (TIReID) facilitated by the breakthrough of cross-modal embedding models, existing methods often struggle to distinguish challenging candidate images due to intrinsic limitations, such as network architecture and data quality. To address these issues, we propose an Interactive Cross-modal Learning framework (ICL), which leverages human-centered interaction to enhance the discriminability of text queries through external multimodal knowledge. To achieve this, we propose a plug-and-play Test-time Humane-centered Interaction (THI) module, which performs visual question answering focused on human characteristics, facilitating multi-round interactions with a multimodal large language model (MLLM) to align query intent with latent target images. Specifically, THI refines user queries based on the MLLM responses to reduce the gap to the best-matching images, thereby boosting ranking accuracy. Additionally, to address the limitation of low-quality training texts, we introduce a novel Reorganization Data Augmentation (RDA) strategy based on information enrichment and diversity enhancement to enhance query discriminability by enriching, decomposing, and reorganizing person descriptions. Extensive experiments on four TIReID benchmarks, i.e., CUHK-PEDES, ICFG-PEDES RSTPReid, RSTPReid, and UFine6926, demonstrate that our method achieves remarkable performance with substantial improvement. Code is available at https://github.com/QinYang79/ICL.

Multi-view clustering (MVC) aims to exploit complementary information from diverse views to enhance clustering performance. Since pseudo-labels can provide additional semantic information, many MVC methods have been proposed to guide unsupervised multi-view learning through pseudo-labels. These methods implicitly assume that the predicted pseudo-labels are predicted correctly. However, due to the challenges in training a flawless unsupervised model, this assumption can be easily violated, thereby leading to the Noisy Pseudo-label Problem (NPP). Moreover, these existing approaches typically rely on the assumption of perfect cross-view alignment. In practice, it is frequently compromised due to noise or sensor differences, thereby resulting in the Noisy Correspondence Problem (NCP). Based on the above observations, we reveal and study unsupervised multi-view learning under NPP and NCP. To this end, we propose Robust Noisy Pseudo-label Learning (ROLL) to prevent the overfitting problem caused by both NPP and NCP. Specifically, we first adopt traditional contrastive learning to warm up the model, thereby generating the pseudo-labels in a self-supervised manner. Afterward, we propose noise-tolerance pseudo-label learning to deal with the noise in the predicted pseudo-labels, thereby embracing the robustness against NPP. To further mitigate the overfitting problem, we present robust multi-view contrastive learning to mitigate the negative impact of NCP. Extensive experiments on five multi-view datasets demonstrate the superior clustering performance of our ROLL compared to 11 state-of-the-art methods.

The success of most existing cross-modal retrieval methods heavily relies on the assumption that the given queries follow the same distribution of the source domain. However, such an assumption is easily violated in real-world scenarios due to the complexity and diversity of queries, thus leading to the query shift problem. Specifically, query shift refers to the online query stream originating from the domain that follows a different distribution with the source one. In this paper, we observe that query shift would not only diminish the uniformity (namely, within-modality scatter) of the query modality but also amplify the gap between query and gallery modalities. Based on the observations, we propose a novel method dubbed Test-time adaptation for Cross-modal Retrieval (TCR). In brief, TCR employs a novel module to refine the query predictions (namely, retrieval results of the query) and a joint objective to prevent query shift from disturbing the common space, thus achieving online adaptation for the cross-modal retrieval models with query shift. Expensive experiments demonstrate the effectiveness of the proposed TCR against query shift. Code is available at https://github.com/XLearning-SCU/2025-ICLR-TCR.

Despite the significant advancements in Text-to-SQL (Text2SQL) facilitated by large language models (LLMs), the latest state-of-the-art techniques are still trapped in the in-context learning of closed-source LLMs (e.g., GPT-4), which limits their applicability in open scenarios. To address this challenge, we propose a novel RObust mUltitask Tuning and collaboration mEthod (ROUTE) to improve the comprehensive capabilities of open-source LLMs for Text2SQL, thereby providing a more practical solution. Our approach begins with multi-task supervised fine-tuning (SFT) using various synthetic training data related to SQL generation. Unlike existing SFT-based Text2SQL methods, we introduced several additional SFT tasks, including schema linking, noise correction, and continuation writing. Engaging in a variety of SQL generation tasks enhances the model's understanding of SQL syntax and improves its ability to generate high-quality SQL queries. Additionally, inspired by the collaborative modes of LLM agents, we introduce a Multitask Collaboration Prompting (MCP) strategy. This strategy leverages collaboration across several SQL-related tasks to reduce hallucinations during SQL generation, thereby maximizing the potential of enhancing Text2SQL performance through explicit multitask capabilities. Extensive experiments and in-depth analyses have been performed on eight open-source LLMs and five widely-used benchmarks. The results demonstrate that our proposal outperforms the latest Text2SQL methods and yields leading performance.

Text-to-visual retrieval often struggles with semantic redundancy and granularity mismatches between textual queries and visual content. Unlike existing methods that address these challenges during training, we propose VISual Abstraction (VISA), a test-time approach that enhances retrieval by transforming visual content into textual descriptions using off-the-shelf large models. The generated text descriptions, with their dense semantics, naturally filter out low-level redundant visual information. To further address granularity issues, VISA incorporates a question-answering process, enhancing the text description with the specific granularity information requested by the user. Extensive experiments demonstrate that VISA brings substantial improvements in text-to-image and text-to-video retrieval for both short- and long-context queries, offering a plug-and-play enhancement to existing retrieval systems.

Multi-view learning methods primarily focus on enhancing decision accuracy but often neglect the uncertainty arising from the intrinsic drawbacks of data, such as noise, conflicts, etc. To address this issue, several trusted multi-view learning approaches based on the Evidential Theory have been proposed to capture uncertainty in multi-view data. However, their performance is highly sensitive to conflicting views, and their uncertainty estimates, which depend on the total evidence and the number of categories, often underestimate uncertainty for conflicting multi-view instances due to the neglect of inherent conflicts between belief masses. To accurately classify conflicting multi-view instances and precisely estimate their intrinsic uncertainty, we present a novel Deep Fuzzy Multi-View Learning (FUML) method. Specifically, FUML leverages Fuzzy Set Theory to model the outputs of a classification neural network as fuzzy memberships, incorporating both possibility and necessity measures to quantify category credibility. A tailored loss function is then proposed to optimize the category credibility. To further enhance uncertainty estimation, we propose an entropy-based uncertainty estimation method leveraging category credibility. Additionally, we develop a Dual Reliable Multi-view Fusion (DRF) strategy that accounts for both view-specific uncertainty and inter-view conflict to mitigate the influence of conflicting views in multi-view fusion. Extensive experiments demonstrate that our FUML achieves state-of-the-art performance in terms of both accuracy and reliability.

Physics-informed neural networks (PINNs) aim to constrain the outputs and gradients of deep learning models to satisfy specified governing physics equations, which have demonstrated significant potential for solving partial differential equations (PDEs). Although existing PINN methods have achieved pleasing performance, they always treat both easy and hard sample points indiscriminately, especially ones in the physical boundaries. This easily causes the PINN model to fall into undesirable local minima and unstable learning, thereby resulting in an Unbalanced Prediction Problem (UPP). To deal with this daunting problem, we propose a novel framework named Cognitive Physical Informed Neural Network (CoPINN) that imitates the human cognitive learning manner from easy to hard. Specifically, we first employ separable subnetworks to encode independent one-dimensional coordinates and apply an aggregation scheme to generate multi-dimensional predicted physical variables. Then, during the training phase, we dynamically evaluate the difficulty of each sample according to the gradient of the PDE residuals. Finally, we propose a cognitive training scheduler to progressively optimize the entire sampling regions from easy to hard, thereby embracing robustness and generalization against predicting physical boundary regions. Extensive experiments demonstrate that our CoPINN achieves state-of-the-art performance, particularly significantly reducing prediction errors in stubborn regions.

Traditional text-based person ReID assumes that person descriptions from witnesses are complete and provided at once. However, in real-world scenarios, such descriptions are often partial or vague. To address this limitation, we introduce a new task called interactive person re-identification (Inter-ReID). Inter-ReID is a dialogue-based retrieval task that iteratively refines initial descriptions through ongoing interactions with the witnesses. To facilitate the study of this new task, we construct a dialogue dataset that incorporates multiple types of questions by decomposing fine-grained attributes of individuals. We further propose LLaVA-ReID, a question model that generates targeted questions based on visual and textual contexts to elicit additional details about the target person. Leveraging a looking-forward strategy, we prioritize the most informative questions as supervision during training. Experimental results on both Inter-ReID and text-based ReID benchmarks demonstrate that LLaVA-ReID significantly outperforms baselines.

Recently, deep unsupervised hashing has gained considerable attention in image retrieval due to its advantages in cost-free data labeling, computational efficiency, and storage savings. Although existing methods achieve promising performance by leveraging inherent visual structures within the data, they primarily focus on learning discriminative features from unlabeled images through limited internal knowledge, resulting in an intrinsic upper bound on their performance. To break through this intrinsic limitation, we propose a novel method, called Deep Unsupervised Hashing with External Guidance (DUH-EG), which incorporates external textual knowledge as semantic guidance to enhance discrete representation learning. Specifically, our DUH-EG: i) selects representative semantic nouns from an external textual database by minimizing their redundancy, then matches images with them to extract more discriminative external features; and ii) presents a novel bidirectional contrastive learning mechanism to maximize agreement between hash codes in internal and external spaces, thereby capturing discrimination from both external and intrinsic structures in Hamming space. Extensive experiments on four benchmark datasets demonstrate that our DUH-EG remarkably outperforms existing state-of-the-art hashing methods.

Multi-view representation learning methods typically follow a consistent-and-specific pipeline that aims at extracting latent representations for an entity from its multiple observable views to facilitate downstream tasks. However, most of them overlook the complex underlying correlation between different views. To solve this issue, we delve into a well-known property of neural networks (NNs) that NNs tend to learn simple patterns first and then hard ones. In our case, view-consistent representations are simple patterns and view-specific representations are hard. To this end, we propose to disentangle view-consistency and view-specificity and learn them gradually. Specifically, we devise a novel curriculum learning approach that adjusts the whole model to learn view-consistent representations first and then progressively view-specific representations. Besides, we saddle each view with a learnable prior that allows each view-specific representation to appropriate its distribution. Moreover, we incorporate a mixture-of-experts layer and a disentangling module to further enhance the quality of the learned representations. Extensive experiments on five real-world datasets show that the proposed model outperforms its counterparts markedly. The code is available at https://github.com/XLearning-SCU/2025-IJCAI-CL2P.

Multimodal learning is pivotal for the advancement of artificial intelligence, enabling machines to integrate complementary information from diverse data sources for holistic perception and understanding. Despite significant progress, existing methods struggle with challenges such as noisy inputs, noisy correspondence, and the inherent uncertainty of multimodal data, limiting their reliability and robustness. To address these issues, this paper presents a novel Probabilistic Multimodal Learning framework (PML) that models each data point as a von Mises-Fisher (vMF) distribution, effectively capturing intrinsic uncertainty and enabling robust fusion. Unlike traditional Gaussian-based models, PML learns directional representation with a concentration parameter to quantify reliability directly, enhancing stability and interpretability. To enhance discrimination, we propose a von Mises-Fisher Prototypical Contrastive Learning paradigm (vMF-PCL), which projects data onto a hypersphere by pulling within-class samples closer to their class prototype while pushing between-class prototypes apart, adaptively learning the reliability estimations. Building upon the estimated reliability, we develop a Reliable Multimodal Fusion mechanism (RMF) that dynamically adjusts the contribution and conflict of each modality, ensuring robustness against noisy data, noisy correspondence, and uncertainty. Extensive experiments on nine benchmarks demonstrate the superiority of PML, consistently outperforming 14 state-of-the-art methods. Code is available at https://github.com/XLearning-SCU/2025-IJCAI-PML.

Most existing multi-view representation learning methods assume view-completeness and noise-free data. However, such assumptions are strong in real-world applications. Despite advances in methods tailored to view-missing or noise problems individually, a one-size-fits-all approach that concurrently addresses both remains unavailable. To this end, we propose a holistic method, called Dual-masked Variational Autoencoders (DualVAE), which aims at learning robust multi-view representation. The DualVAE exhibits an innovative amalgamation of dual-masked prediction, mixture-of-experts learning, representation disentangling, and a joint loss function in wrapping up all components. The key novelty lies in the dual-masked (view-mask and patch-mask) mechanism to mimic missing views and noisy data. Extensive experiments on four multi-view datasets show the effectiveness of the proposed method and its superior performance in comparison to baselines. The code is available at https://github.com/XLearning-SCU/2025-IJCAI-DualVAE.

Text-3D Scene Retrieval (T3SR) aims to retrieve relevant scenes using linguistic queries. Although traditional T3SR methods have made significant progress in capturing fine-grained associations, they implicitly assume that query descriptions are information-complete. In practical deployments, however, limited by the capabilities of users and models, it is difficult or even impossible to directly obtain a perfect textual query suiting the entire scene and model, thereby leading to performance degradation. To address this issue, we propose a novel Interactive Text-3D Scene Retrieval Method (IDeal), which promotes the enhancement of the alignment between texts and 3D scenes through continuous interaction. To achieve this, we present an Interactive Retrieval Refinement framework (IRR), which employs a questioner to pose contextually relevant questions to an answerer in successive rounds that either promote detailed probing or encourage exploratory divergence within scenes. Upon the iterative responses received from the answerer, IRR adopts a retriever to perform both feature-level and semantic-level information fusion, facilitating scene-level interaction and understanding for more precise re-rankings. To bridge the domain gap between queries and interactive texts, we propose an Interaction Adaptation Tuning strategy (IAT). IAT mitigates the discriminability and diversity risks among augmented text features that approximate the interaction text domain, achieving contrastive domain adaptation for our retriever. Extensive experimental results on three datasets demonstrate the superiority of IDeal. Code is available at https://github.com/Yangl1nFeng/IDeal.

Grounding target objects in 3D environments via natural language is a fundamental capability for autonomous agents to successfully fulfill user requests. Almost all existing works typically assume that the target object lies within a known scene and focus solely on in-scene localization. In practice, however, agents often encounter unknown or previously visited environments and need to search across a large archive of scenes to ground the described object, thereby invalidating this assumption. To address this, we reveal a novel task called Cross-Scene Spatial Reasoning and Grounding (CSSRG), which aims to locate a described object anywhere across an entire collection of 3D scenes rather than predetermined scenes. Due to the difference from existing 3D visual grounding, CSSRG poses two challenges: the prohibitive cost of exhaustively traversing all scenes and more complex cross-modal spatial alignment. To address the challenges, we propose a Cross-Scene 3D Object Reasoning Framework (CoRe), which adopts a matching-then-grounding pipeline to reduce computational overhead. Specifically, CoRe consists of i) a Robust Text-Scene Aligning (RTSA) module that learns global scene representations for robust alignment between object descriptions and the corresponding 3D scenes, enabling efficient retrieval of candidate scenes; and ii) a Tailored Word-Object Associating (TWOA) module that establishes fine-grained alignment between words and target objects to filter out redundant context, supporting precise object-level reasoning and alignment. Additionally, to benchmark CSSRG, we construct a new CrossScene-RETR dataset and evaluation protocol tailored for cross-scene grounding. Extensive experiments across four multimodal datasets demonstrate that CoRe dramatically reduces computational overhead while showing superiority in both scene retrieval and object grounding. Code is available at https://github.com/Yangl1nFeng/CoRe.

In this paper, we investigate source-free domain adaptation (SFDA) for visible-infrared person re-identification (VI-ReID), aiming to adapt a pre-trained source model to an unlabeled target domain without access to source data. To address this challenging setting, we propose a novel learning paradigm, termed Source-Free Visible-Infrared Person Re-Identification (SVIP), which fully exploits the prior knowledge embedded in the source model to guide target domain adaptation. The proposed framework comprises three key components specifically designed for the source-free scenario: 1) a Source-Guided Contrastive Learning (SGCL) module, which leverages the discriminative feature space of the frozen source model as a reference to perform contrastive learning on the unlabeled target data, thereby preserving discrimination without requiring source samples; 2) a Residual Transfer Learning (RTL) module, which learns residual mappings to adapt the target model’s representations while maintaining the knowledge from the source model; and 3) a Structural Consistency-Guided Cross-modal Alignment (SCCA) module, which enforces reciprocal structural constraints between visible and infrared modalities to identify reliable cross-modal pairs and achieve robust modality alignment without source supervision. Extensive experiments on benchmark datasets demonstrate that SVIP substantially enhances target domain performance and outperforms existing unsupervised VI-ReID methods under source-free settings.

Conventional representation learning methods learn a universal representation that primarily captures dominant semantics, which may not always align with customized downstream tasks. For instance, in animal habitat analysis, researchers prioritize scene-related features, whereas universal embeddings emphasize categorical semantics, leading to suboptimal results. As a solution, existing approaches resort to supervised fine-tuning, which however incurs high computational and annotation costs. In this paper, we propose Conditional Representation Learning (CRL), aiming to extract representations tailored to arbitrary user-specified criteria. Specifically, we reveal that the semantics of a space are determined by its basis, thereby enabling a set of descriptive words to approximate the basis for a customized feature space. Building upon this insight, given a user-specified criterion, CRL first employs a large language model (LLM) to generate descriptive texts to construct the semantic basis, then projects the image representation into this conditional feature space leveraging a vision-language model (VLM). The conditional representation better captures semantics for the specific criterion, which could be utilized for multiple customized tasks. Extensive experiments on classification and retrieval tasks demonstrate the superiority and generality of the proposed CRL. The code is available at https://github.com/XLearning-SCU/2025-NeurIPS-CRL.

Vision-Language Pre-training (VLP) has shown promising performance in various tasks by learning a generic image-text representation space. However, most existing VLP methods encounter the Noisy Correspondence (NC) problem which refers to wrongly matched image-text pairs harvested from the wild. In this paper, we empirically study the influence of NC on the VLP model and obtain the following two observations. First, the NC will largely degrade the performance in downstream tasks even via fine-tuning, indicating the necessity of handling NC in the pre-training period. Second, the influence of NC varies in different pre-training objectives, suggesting the objective-customized solution for achieving NC robustness. Based on the above observations, we propose a novel NoisE-robust Vision-languagE pRe-training method (NEVER) to endow the VLP model with robustness against NC. In brief, NEVER first divides the training data into clean and noisy subsets in a progressive and adaptive manner. Then NEVER employs the positive learning (PL) and negative learning (NL) on the splits to enjoy model convergence and noise robustness, respectively. To further handle the false negative in PL and NL, NEVER proposes to smoothen and sharpen the training targets with the predictions from a twin momentum model. Extensive experiments on the various V+L tasks verify the effectiveness of the proposed method.

In this paper, we study two challenging but less-touched problems in image restoration, namely, i) how to quantify the relationship between image degradations and ii) how to improve the performance of a specific restoration task using the quantified relationship. To tackle the first challenge, we propose the Degradation Relationship Index (DRI), which is defined as the mean drop rate difference in validation loss between two models, where one trained solely with anchor degradation and the other trained with both anchor and auxiliary degradations. By quantifying degradation relationship using DRI, we reveal that i) a positive DRI consistently indicates performance improvement when a beneficial auxiliary degradation is incorporated during training; ii) the proportion of auxiliary degradation is crucial to the anchor task performance. In other words, performance improvement is achieved only when the anchor and auxiliary degradations are combined in an appropriate proportion. Based on these observations, we further propose a simple yet effective Degradation Proportion Determination (DPD) method to estimate whether a given degradation combinations can enhance performance on the anchor restoration task with the assistance of auxiliary degradation. Extensive experimental results verify the effectiveness and generalizability of our method on noise, rain streak, haze and snow.

Unsupervised visible-infrared person re-identification (UVI-ReID) aims at retrieving pedestrian images of the same individual across distinct modalities, presenting challenges due to the inherent heterogeneity gap and the absence of cost-prohibitive annotations. Although existing methods employ self-training with clustering-generated pseudo-labels to bridge this gap, they always implicitly assume that these pseudo-labels are predicted correctly. In practice, however, this presumption is impossible to satisfy due to the difficulty of training a perfect model let alone without any ground truths, resulting in pseudo-labeling errors. Based on the observation, this study introduces a new learning paradigm for UVI-ReID considering Pseudo-Label Noise (PLN), which encompasses three challenges: noise overfitting, error accumulation, and noisy cluster correspondence. To conquer these challenges, we propose a novel robust duality learning framework (RoDE) for UVI-ReID to mitigate the adverse impact of noisy pseudo-labels. Specifically, for noise overfitting, we propose a novel Robust Adaptive Learning mechanism (RAL) to dynamically prioritize clean samples while deprioritizing noisy ones, thus avoiding overemphasizing noise. To circumvent error accumulation of self-training, where the model tends to confirm its mistakes, RoDE alternately trains dual distinct models using pseudo-labels predicted by their counterparts, thereby maintaining diversity and avoiding collapse into noise. However, this will lead to cross-cluster misalignment between the two distinct models, not to mention the misalignment between different modalities, resulting in dual noisy cluster correspondence and thus difficult to optimize. To address this issue, a Cluster Consistency Matching mechanism (CCM) is presented to ensure reliable alignment across distinct modalities as well as across different models by leveraging cross-cluster similarities. Extensive experiments on three benchmark datasets demonstrate the effectiveness of the proposed RoDE.

Visual-textual retrieval, as a link between computer vision and natural language processing, aims at jointly learning visual-semantic relevance to bridge the heterogeneity gap across visual and textual spaces. Existing methods conduct retrieval only relying on the ranking of pairwise similarities, but they cannot self-evaluate the uncertainty of retrieved results, resulting in unreliable retrieval and hindering interpretability. To address this problem, we propose a novel Trust-Consistent Learning framework (TCL) to endow visual-textual retrieval with uncertainty evaluation for trustworthy retrieval. More specifically, TCL first models the matching evidence according to cross-modal similarity to estimate the uncertainty for cross-modal uncertainty-aware learning. Second, a simple yet effective consistency module is presented to enforce the subjective opinions of bidirectional learning to be consistent for high reliability and accuracy. Finally, extensive experiments are conducted to demonstrate the superiority and generalizability of TCL on six widely-used benchmark datasets, i.e., Flickr30K, MS-COCO, MSVD, MSR-VTT, ActivityNet, and DiDeMo. Furthermore, some qualitative experiments are carried out to provide comprehensive and insightful analyses for trustworthy visual-textual retrieval, verifying the reliability and interoperability of TCL. The code is available in https://github.com/QinYang79/TCL.

Outlier Detection (OD) has attracted extensive research due to its application in many fields. The idea of neighborhood computing is one of the widely used methods in outlier analysis. Nevertheless, these methods mainly use certainty strategies to model outlier detection, so they cannot effectively handle the fuzzy information in the dataset. Moreover, they mainly focus on dealing with outlier detection in numerical data and cannot effectively find outliers in mixed-attribute data. Fuzzy information granulation theory is an effective granular computing model that allows objects to belong to a set to a certain extent (i.e., membership degree), which makes it possible to better handle uncertainty problems such as fuzziness. In this work, we propose an outlier detection model based on fuzzy neighborhoods. First, a hybrid fuzzy similarity is constructed to granulate the set of objects to form fuzzy information granules. Second, the fuzzy$k$-nearest neighbor is defined to describe the fuzzy local information. Then, the fuzzy neighborhood density is defined to indicate the degree of aggregation of each object. The smaller the fuzzy neighborhood density of an object, the more likely it is to be an outlier. Based on this idea, the fuzzy neighborhood deviation degree is defined to quantify the degree of outliers of objects. Finally, the fuzzy deviation degree on the set of conditional attributes is constructed to indicate the outlier scores of objects. Experimental comparisons with state-of-the-art methods show that the proposed method has a significant improvement on the AUC index and applies to three types of data.

In this paper, we present and study a new instance-level retrieval task: PointCloud-Text Matching (PTM), which aims to identify the exact cross-modal instance that matches a given point-cloud query or text query. PTM has potential applications in various scenarios, such as indoor/urban-canyon localization and scene retrieval. However, there is a lack of suitable and targeted datasets for PTM in practice. To address this issue, we present a new PTM benchmark dataset, namely SceneDepict-3D2T. We observe that the data poses significant challenges due to its inherent characteristics, such as the sparsity, noise, or disorder of point clouds and the ambiguity, vagueness, or incompleteness of texts, which render existing cross-modal matching methods ineffective for PTM. To overcome these challenges, we propose a PTM baseline, namedRobust PointCloud-TextMatching method (RoMa). RoMa consists of two key modules: a Dual Attention Perception module (DAP) and a Robust Negative Contrastive Learning module (RNCL). Specifically, DAP leverages token-level and feature-level attention mechanisms to adaptively focus on useful local and global features, and aggregate them into common representations, thereby reducing the adverse impact of noise and ambiguity. To handle noisy correspondence, RNCL enhances robustness against mismatching by dividing negative pairs into clean and noisy subsets and assigning them forward and reverse optimization directions, respectively. We conduct extensive experiments on our benchmarks and demonstrate the superiority of our RoMa.

Cross-modal hashing~(CMH) is an efficient technique to retrieve relevant data across different modalities, such as images, texts, and videos, which has attracted more and more attention due to its low storage cost and fast query speed. Although existing CMH methods achieve remarkable processes, almost all of them treat all samples of varying difficulty levels without discrimination, thus leaving them vulnerable to noise or outliers. Based on this observation, we reveal and study dual difficulty levels implied in cross-modal hashing learning, \ie instance-level and feature-level difficulty. To address this problem, we propose a novel Dual Self-Paced Cross-Modal Hashing (DSCMH) that mimics human cognitive learning to learn hashing from ``easy'' to ``hard'' in both instance and feature levels, thereby embracing robustness against noise/outliers. Specifically, our DSCMH assigns weights to each instance and feature to measure their difficulty or reliability, and then uses these weights to automatically filter out the noisy and irrelevant data points in the original space. By gradually increasing the weights during training, our method can focus on more instances and features from ``easy'' to ``hard'' in training, thus mitigating the adverse effects of noise or outliers. Extensive experiments are conducted on three widely-used benchmark datasets to demonstrate the effectiveness and robustness of the proposed DSCMH over 12 state-of-the-art CMH methods.

Video-based facial analysis is important for autonomous agents to understand human expressions and sentiments. However, limited labeled data is available to learn effective facial representations. This paper proposes a novel self-supervised face-centric pretraining framework, called PrefAce, which learns transferable video facial representation without labels. The self-supervised learning is performed with an effective landmark-guided global-local tube distillation. Meanwhile, a novel instance-wise update FaceFeat Cache is built to enforce more discriminative and diverse representations for downstream tasks. Extensive experiments demonstrate that the proposed framework learns universal instance-aware facial representations with fine-grained landmark details from videos. The point is that it can transfer across various facial analysis tasks, e.g., Facial Attribute Recognition (FAR), Facial Expression Recognition (FER), DeepFake Detection (DFD), and Lip Synchronization (LS). Our framework also outperforms the state-of-the-art on various downstream tasks, even in low data regimes. Code is available at https://github.com/siyuan-h/PrefAce.

In recent, some robust contrastive multi-view clustering (MvC) methods have been proposed, which construct data pairs from neighborhoods to alleviate the false negative issue, i.e., some intra-cluster samples are wrongly treated as negative pairs. Although promising performance has been achieved by these methods, the false negative issue is still far from addressed and the false positive issue emerges because all in- and out-of-neighborhood samples are simply treated as positive and negative, respectively. To address the issues, we propose a novel robust method, dubbed decoupled contrastive multi-view clustering with high-order random walks (DIVIDE). In brief, DIVIDE leverages random walks to progressively identify data pairs in a global instead of local manner. As a result, DIVIDE could identify in-neighborhood negatives and out-of-neighborhood positives. Moreover, DIVIDE embraces a novel MvC architecture to perform inter- and intra-view contrastive learning in different embedding spaces, thus boosting clustering performance and embracing the robustness against missing views. To verify the efficacy of DIVIDE, we carry out extensive experiments on four benchmark datasets comparing with nine state-of-the-art MvC methods in both complete and incomplete MvC settings. The code is released on https://github.com/XLearning-SCU/2024-AAAI-DIVIDE.

Text-to-image person re-identification (TIReID) is a compelling topic in the cross-modal community, which aims to retrieve the target person based on a textual query. Although numerous TIReID methods have been proposed and achieved promising performance, they implicitly assume the training image-text pairs are correctly aligned, which is not always the case in real-world scenarios. In practice, the image-text pairs inevitably exist under-correlated or even false-correlated, a.k.a noisy correspondence (NC), due to the low quality of the images and annotation errors. To address this problem, we propose a novel Robust Dual Embedding method (RDE) that can learn robust visual-semantic associations even with NC. Specifically, RDE consists of two main components: 1) A Confident Consensus Division (CCD) module that leverages the dual-grained decisions of dual embedding modules to obtain a consensus set of clean training data, which enables the model to learn correct and reliable visual-semantic associations. 2) A Triplet Alignment Loss (TAL) relaxes the conventional Triplet Ranking loss with the hardest negative samples to a log-exponential upper bound over all negative ones, thus preventing the model collapse under NC and can also focus on hard-negative samples for promising performance. We conduct extensive experiments on three public benchmarks, namely CUHK-PEDES, ICFG-PEDES, and RSTPReID, to evaluate the performance and robustness of our RDE. Our method achieves state-of-the-art results both with and without synthetic noisy correspondences on all three datasets. Code is available at https://github.com/QinYang79/RDE.

Test-time adaptation (TTA) has emerged as a new paradigm for reconciling distribution shifts across domains without accessing source data. However, existing TTA methods mainly concentrate on uni-modal tasks, overlooking the complexity of multi-modal scenarios. In this paper, we delve into the multi-modal test-time adaptation and reveal a new challenge named reliability bias. Different from the definition of traditional distribution shifts, reliability bias refers to the information discrepancies across different modalities derived from intra-modal distribution shifts. To solve the challenge, we propose a novel method, dubbed REliable fusion and robust ADaptation (READ). On the one hand, unlike the existing TTA paradigm that mainly repurposes the normalization layers, READ employs a new paradigm that modulates the attention between modalities in a self-adaptive way, supporting reliable fusion against reliability bias. On the other hand, READ adopts a novel objective function for robust multi-modal adaptation, where the contributions of confident predictions could be amplified and the negative impacts of noisy predictions could be mitigated. Moreover, we introduce two new benchmarks to facilitate comprehensive evaluations of multi-modal TTA under reliability bias. Extensive experiments on the benchmarks verify the effectiveness of our method against multi-modal reliability bias. The code and benchmarks are available at https://github.com/XLearning-SCU/2024-ICLR-READ.

The core of clustering lies in incorporating prior knowledge to construct supervision signals. From classic k-means based on data compactness to recent contrastive clustering guided by self-supervision, the evolution of clustering methods intrinsically corresponds to the progression of supervision signals. At present, substantial efforts have been devoted to mining internal supervision signals from data. Nevertheless, the abundant external knowledge such as semantic descriptions, which naturally conduces to clustering, is regrettably overlooked. In this work, we propose leveraging external knowledge as a new supervision signal to guide clustering. To implement and validate our idea, we design an externally guided clustering method (Text-Aided Clustering, TAC), which leverages the textual semantics of WordNet to facilitate image clustering. Specifically, TAC first selects and retrieves WordNet nouns that best distinguish images to enhance the feature discriminability. Then, TAC collaborates text and image modalities by mutually distilling cross-modal neighborhood information. Experiments demonstrate that TAC achieves state-of-the-art performance on five widely used and three more challenging image clustering benchmarks, including the full ImageNet-1K dataset. The code can be accessed at https://github.com/XLearning-SCU/2024-ICML-TAC.

Although multi-view learning has achieved remarkable progress over the past decades, most existing methods implicitly assume that all views (or modalities) are well-aligned. In practice, however, collecting fully aligned views is challenging due to complexities and discordances in time and space, resulting in the Partially View-unaligned Problem (PVP), such as audio-video asynchrony caused by network congestion. While some methods are proposed to align the unaligned views by learning view-invariant representations, almost all of them overlook specific information across different views for complementarity, limiting performance improvement. To address these problems, we propose a robust framework, dubbed VariatIonal ConTrAstive Learning (VITAL), designed to learn both common and specific information simultaneously. To be specific, each data sample is first modeled as a Gaussian distribution in the latent space, where the mean estimates the most probable common information, while the variance indicates view-specific information. Second, by using variational inference, VITAL conducts intra- and inter-view contrastive learning to preserve common and specific semantics in the distribution representations, thereby achieving comprehensive perception. As a result, the common representation (mean) could be used to guide category-level realignment, while the specific representation (variance) complements sample semantic information, thereby boosting overall performance. Finally, considering the abundance of False Negative Pairs (FNPs) generated by unsupervised contrastive learning, we propose a robust loss function that seamlessly incorporates FNP rectification into the contrastive learning paradigm. Empirical evaluations on eight benchmark datasets reveal that VITAL outperforms ten state-of-the-art deep clustering baselines, demonstrating its efficacy in both partially and fully aligned scenarios. The Code is available at https://github.com/He-Changhao/2024-MM-VITAL.

Cross-modal hashing has emerged as a promising technique for retrieving relevant information across distinct media types thanks to its low storage cost and high retrieval efficiency. However, the success of most existing methods heavily relies on large-scale well-annotated datasets, which are costly and scarce in the real world due to ubiquitous labeling noise. To tackle this problem, in this paper, we propose a novel framework, termed Noise Resistance Cross-modal Hashing (NRCH), to learn hashing with noisy labels by overcoming two key challenges, i.e. noise overfitting and error accumulation. Specifically, i) to mitigate the overfitting issue caused by noisy labels, we present a novel Robust Contrastive Hashing loss (RCH) to target homologous pairs instead of noisy positive pairs, thus avoiding overemphasizing noise. In other words, RCH enforces the model focus on more reliable positives instead of unreliable ones constructed by noisy labels, thereby enhancing the robustness of the model against noise; ii) to circumvent error accumulation, a Dynamic Noise Separator (DNS) is proposed to dynamically and accurately separate the clean and noisy samples by adaptively fitting the loss distribution, thus alleviate the adverse influence of noise on iterative training. Finally, we conduct extensive experiments on four widely used benchmarks to demonstrate the robustness of our NRCH against noisy labels for cross-modal retrieval. The code is available at: https://github.com/LonganWANG-cs/NRCH.git.

Recently, contrastive multi-view clustering (MvC) has emerged as a promising avenue for analyzing data from heterogeneous sources, typically leveraging the off-the-shelf instances as positives and randomly sampled ones as negatives. In practice, however, this paradigm would unavoidably suffer from the Dual Noisy Correspondence (DNC) problem, where noise compromises the constructions of both positive and negative pairs. Specifically, the complexity of data collection and transmission might mistake some unassociated pairs as positive (namely, false positive correspondence), while the intrinsic one-to-many contrast nature of contrastive MvC would sample some intra-cluster samples as negative (namely, false negative correspondence). To handle this daunting problem, we propose a novel method, dubbed Contextually-spectral based correspondence refinery (CANDY). CANDY dexterously exploits inter-view similarities as \textit{context} to uncover false negatives. Furthermore, it employs a spectral-based module to denoise correspondence, alleviating the negative influence of false positives. Extensive experiments on five widely-used multi-view benchmarks, in comparison with eight competitive multi-view clustering methods, verify the effectiveness of our method in addressing the DNC problem. The code is available at https://github.com/XLearning-SCU/2024-NeurIPS-CANDY.

In the absence of class priors, recent deep clustering methods resort to data augmentation and pseudo-labeling strategies to generate supervision signals. Though achieved remarkable success, existing works struggle to discriminate hard samples at cluster boundaries, mining which is particularly challenging due to their unreliable cluster assignments. To break such a performance bottleneck, we propose incorporating user interaction to facilitate clustering instead of exhaustively mining semantics from the data itself. To be exact, we present Interactive Deep Clustering (IDC), a plug-and-play method designed to boost the performance of pre-trained clustering models with minimal interaction overhead. More specifically, IDC first quantitatively evaluates sample values based on hardness, representativeness, and diversity, where the representativeness avoids selecting outliers and the diversity prevents the selected samples from collapsing into a small number of clusters. IDC then queries the cluster affiliations of high-value samples in a user-friendly manner. Finally, it utilizes the user feedback to finetune the pre-trained clustering model. Extensive experiments demonstrate that IDC could remarkably improve the performance of various pre-trained clustering models, at the expense of low user interaction costs. The code could be accessed at pengxi.me.

Traditional video restoration approaches were designed to recover clean videos from a specific type of degradation, making them ineffective in handling multiple unknown types of degradation. To address this issue, several studies have been conducted and have shown promising results. However, these studies overlook that the degradations in video usually change over time, dubbed time-varying unknown degradations (TUD). To tackle such a less-touched challenge, we propose an innovative method, termed as All-in-one VidEo Restoration Network (AverNet), which comprises two core modules, i.e., Prompt-Guided Alignment (PGA) module and Prompt-Conditioned Enhancement (PCE) module. Specifically, PGA addresses the issue of pixel shifts caused by time-varying degradations by learning and utilizing prompts to align video frames at the pixel level. To handle multiple unknown degradations, PCE recasts it into a conditional restoration problem by implicitly establishing a conditional map between degradations and ground truths. Thanks to the collaboration between PGA and PCE modules, AverNet empirically demonstrates its effectiveness in recovering videos from TUD. Extensive experiments are carried out on two synthesized datasets featuring seven types of degradations with random corruption levels. The code is available at https://github.com/XLearning-SCU/2024-NeurIPS-AverNet.

Robust multi-view learning with incomplete information has received significant attention due to issues such as incomplete correspondences and incomplete instances that commonly affect real-world multi-view applications. Existing approaches heavily rely on paired samples to realign or impute defective ones, but such preconditions cannot always be satisfied in practice due to the complexity of data collection and transmission. To address this problem, we present a novel framework called SeMantic Invariance LEarning (SMILE) for multi-view clustering with incomplete information that does not require any paired samples. To be specific, we discover the existence of invariant semantic distribution across different views, which enables SMILE to alleviate the cross-view discrepancy to learn consensus semantics without requiring any paired samples. The resulting consensus semantics remains unaffected by cross-view distribution shifts, making them useful for realigning/imputing defective instances and forming clusters. We demonstrate the effectiveness of SMILE through extensive comparison experiments with 13 state-of-the-art baselines on five benchmarks. Our approach improves the clustering accuracy of NoisyMNIST from 19.3%/23.2% to 82.7%/69.0% when the correspondences/instances are fully incomplete. We will release the code after acceptance.

The inherent vulnerability of deep learning to adversarial examples poses a significant security challenge. Although existing defense methods have partially mitigated the harm caused by adversarial attacks, they are still unable to meet practical needs due to their high cost, high latency, and poor defense performance. In this paper, we propose an advanced plug-and-play adversarial purification model called DifFilter. Specifically, we use the superior generative properties of diffusion models to denoise adversarial perturbations and recover clean images. To make Gaussian noise disrupt adversarial perturbations while preserving the real semantic information in the input image, we extend forward diffusion to an infinite number of noise scales so that the distribution of perturbation data evolves with increasing noise according to stochastic differential equations. In the inverse denoising process, we develop a score-based model learning method to restore the input prior distribution to the data distribution of the original clean sample, resulting in stronger purification effects. Additionally, we propose an efficient sampling method to accelerate the computation speed of inverse process, greatly reducing the time cost of purification. We conduct extensive experiments to evaluate the defense generalization performance of DifFilter. The results demonstrate that our method not only surpasses existing defense methods in defense robustness under strong adaptive and black-box attacks but also achieves higher certificate accuracy than the baseline. Furthermore, DifFilter can be combined with adversarial training to further improve defense robustness.

This paper presents a novel method for supervised multi-view representation learning, which projects multiple views into a latent common space while preserving the discrimination and intrinsic structure of each view. Specifically, an apriori discriminant similarity graph is first constructed based on labels and pairwise relationships of multi-view inputs. Then, view-specific networks progressively map inputs to common representations whose affinity approximates the constructed graph. To achieve graph consistency, discrimination, and cross-view invariance, the similarity graph is enforced to meet the following constraints: 1) pairwise relationship should be consistent between the input space and common space for each view; 2) within-class similarity is larger than any between-class similarity for each view; 3) the inter-view samples from the same (or different) classes are mutually similar (or dissimilar). Consequently, the intrinsic structure and discrimination are preserved in the latent common space using an apriori approximation schema. Moreover, we present a sampling strategy to approach a sub-graph sampled from the whole similarity structure instead of approximating the graph of the whole dataset explicitly, thus benefiting lower space complexity and the capability of handling large-scale multi-view datasets. Extensive experiments show the promising performance of our method on five datasets by comparing it with 18 state-of-the-art methods.

The rise of the metaverse and the increasing volume of heterogeneous 2D and 3D data have created a growing demand for cross-modal retrieval, enabling users to query semantically relevant data across different modalities. Existing methods heavily rely on class labels to bridge semantic correlations; however, collecting large-scale, well-labeled data is expensive and often impractical, making unsupervised learning more attractive and feasible. Nonetheless, unsupervised cross-modal learning faces challenges in bridging semantic correlations due to the lack of label information, leading to unreliable discrimination. In this paper, we reveal and study a novel problem: unsupervised cross-modal learning with noisy pseudo-labels. To address this issue, we propose a 2D-3D unsupervised multimodal learning framework that leverages multimodal data. Our framework consists of three key components: 1) Self-matching Supervision Mechanism (SSM) warms up the model to encapsulate discrimination into the representations in a self-supervised learning manner. 2) Robust Discriminative Learning (RDL) further mines the discrimination from the learned imperfect predictions after warming up. To tackle the noise in the predicted pseudo labels, RDL leverages a novel Robust Concentrating Learning Loss (RCLL) to alleviate the influence of the uncertain samples, thus embracing robustness against noisy pseudo labels. 3) Modality-invariance Learning Mechanism (MLM) minimizes the cross-modal discrepancy to enforce SSM and RDL to produce common representations. We conduct comprehensive experiments on four 2D-3D multimodal datasets, comparing our method against 14 state-of-the-art approaches, thereby demonstrating its effectiveness and superiority.

The success of existing cross-modal retrieval (CMR) methods heavily rely on the assumption that the annotated cross-modal correspondence is faultless. In practice, however, the correspondence of some pairs would be inevitably contaminated during data collection or annotation, thus leading to the so-called Noisy Correspondence (NC) problem. To alleviate the influence of NC, we propose a novel method termed Consistency REfining And Mining (CREAM) by revealing and exploiting the difference between correspondence and consistency. Specifically, the correspondence and the consistency only be coincident for true positive and true negative pairs, while being distinct for false positive and false negative pairs. Based on the observation, CREAM employs a collaborative learning paradigm to detect and rectify the correspondence of positives, and a negative mining approach to explore and utilize the consistency. Thanks to the consistency refining and mining strategy of CREAM, the overfitting on the false positives could be prevented and the consistency rooted in the false negatives could be exploited, thus leading to a robust CMR method. Extensive experiments verify the effectiveness of our method on three image-text benchmarks including Flickr30K, MS-COCO, and Conceptual Captions. Furthermore, we adopt our method into the graph matching task and the results demonstrate the robustness of our method against fine-grained NC problem. The code is available on https://github.com/XLearning-SCU/2024-TIP-CREAM.

Deep multi-view clustering leverages deep neural networks to achieve promising performance, but almost all existing methods implicitly assume that all views are aligned correctly. This assumption is unrealistic in many real-world scenarios, where noise, occlusion, or sensor differences can inevitably cause misaligned data. Based on this observation, we reveal and study a practical but understudied problem in multi-view clustering (MVC), i.e., noisy correspondence (NC). Considering this problem, we argue that the main challenge is to prevent the model from overfiting NC. To this end, we propose a novel Robust Multi-view Clustering with Noisy Correspondence (RMCNC) method, which alleviates the influence of the misaligned pairs from multi-view data. To be specific, we first compute a united probability with all positive pairs to learn cross-view alignment consistency, thereby alleviating the adverse impact of the individual false positives. To further mitigate the overfitting problem, we propose a noise-tolerance multi-view contrastive loss that avoids overemphasizing noisy data. Moreover, RMCNC is a unified framework, which can deal with both partially view-aligned and NC problems in multi-view clustering. To the best of our knowledge, it could be the first study on NC in multi-view clustering. The experimental results on eight benchmark datasets indicate our RMCNC achieves competitive performance and robustness.

In recent years, image hashing has attracted more and more attention in practical retrieval applications due to its low storage cost and high query speed. Although existing hashing methods have achieved promising performance, they always treat both easy and hard points without discrimination, thus easily getting stuck into bad local minima, especially in the presence of noise or outliers. In this paper, we reveal that there exist dual difficulty levels hindering binarization in learning to hash, i.e., samples and bits. To overcome this problem, we propose a novel Dual Self-Paced Hashing method (DSPH) for image retrieval, which learns binary codes by not only evolving from ‘easy’ to ‘hard’ samples but also from ‘easy’ to ‘hard’ bits, mimicking the cognitive learning process from easy to difficult. Specifically, we endow each sample and bit with a weight to estimate the reliability/ ease of the row and column, respectively. Then both sampleand bit-level weighting are conducted on rows and columns of Hamming space to enforce the model focus on reliable/easy examples and bits. By gradually increasing the weights during model optimization, more samples and bits are automatically involved in training from ‘easy’ to ‘hard’ via our dual self-paced method, thereby alleviating the adverse impact caused by noises or outliers to learn robust hashing models. Extensive experiments are conducted on four benchmark datasets to demonstrate the superiority and robustness of the proposed DSPH. Our source code has been released athttps://github.com/sunyuan-cs/DSPH.

Cross-modal hashing (CMH) has gained much attention due to its effectiveness and efficiency in facilitating efficient retrieval between different modalities. Whereas, most existing methods unconsciously ignore the hierarchical structural information of the data, and often learn a single-layer hash function to directly transform cross-modal data into common low-dimensional hash codes in one step. This sudden drop of dimension and the huge semantic gap can cause the discriminative information loss. To this end, we adopt a coarse-to-fine progressive mechanism and propose a novelHierarchical Consensus Cross-Modal Hashing (HCCH). Specifically, to mitigate the loss of important discriminative information, we propose a coarse-to-fine hierarchical hashing scheme that utilizes a two-layer hash function to refine the beneficial discriminative information gradually. And then, the$\ell _{2,1}$-norm is imposed on the layer-wise hash function to alleviate the effects of redundant and corrupted features. Finally, we present consensus learning to effectively encode data into a consensus space in such a progressive way, thereby reducing the semantic gap progressively. Through extensive contrast experiments with some advanced CMH methods, the effectiveness and efficiency of our HCCH method are demonstrated on four benchmark datasets.

A variety of single-cell RNA-seq (scRNA-seq) clustering methods has achieved great success in discovering cellular phenotypes. However, it remains challenging when the data confounds with batch effects brought by different experimental conditions or technologies. Namely, the data partitions would be biased toward these nonbiological factors. Meanwhile, the batch differences are not always much smaller than true biological variations, hindering the cooperation of batch integration and clustering methods. To overcome this challenge, we propose single-cell RNA-seq debiased clustering (SCDC), an end-to-end clustering method that is debiased toward batch effects by disentangling the biological and nonbiological information from scRNA-seq data during data partitioning. In six analyses, SCDC qualitatively and quantitatively outperforms both the state-of-the-art clustering and batch integration methods in handling scRNA-seq data with batch effects. Furthermore, SCDC clusters data with a linearly increasing running time with respect to cell numbers and a fixed graphics processing unit (GPU) memory consumption, making it scalable to large datasets. The code will be released on Github.

Cross-domain image retrieval aims at retrieving images across different domains to excavate cross-domain classificatory or correspondence relationships. This paper studies a less-touched problem of cross-domain image retrieval, i.e., unsupervised cross-domain image retrieval, considering the following practical assumptions: (i) no correspondence relationship, and (ii) no category annotations. It is challenging to align and bridge distinct domains without cross-domain correspondence. To tackle the challenge, we present a novel Correspondence-free Domain Alignment (CoDA) method to effectively eliminate the cross-domain gap through In-domain Self-matching Supervision (ISS) and Cross-domain Classifier Alignment (CCA). To be specific, ISS is presented to encapsulate discriminative information into the latent common space by elaborating a novel self-matching supervision mechanism. To alleviate the cross-domain discrepancy, CCA is proposed to align distinct domain-specific classifiers. Thanks to the ISS and CCA, our method could encode the discrimination into the domain-invariant embedding space for unsupervised cross-domain image retrieval. To verify the effectiveness of the proposed method, extensive experiments are conducted on four benchmark datasets compared with six state-of-the-art methods.

Recently, with the advent of Metaverse and AI Generated Content, cross-modal retrieval becomes popular with a burst of 2D and 3D data. However, this problem is challenging given the heterogeneous structure and semantic discrepancies. Moreover, imperfect annotations are ubiquitous given the ambiguous 2D and 3D content, thus inevitably producing noisy labels to degrade the learning performance. To tackle the problem, this paper proposes a robust 2D-3D retrieval framework (RONO) to robustly learn from noisy multimodal data. Specifically, one novel Robust Discriminative Center Learning mechanism (RDCL) is proposed in RONO to adaptively distinguish clean and noisy samples for respectively providing them with positive and negative optimization directions, thus mitigating the negative impact of noisy labels. Besides, we present a Shared Space Consistency Learning mechanism (SSCL) to capture the intrinsic information inside the noisy data by minimizing the cross-modal and semantic discrepancy between common space and label space simultaneously. Comprehensive mathematical analyses are given to theoretically prove the noise tolerance of the proposed method. Furthermore, we conduct extensive experiments on four 3D-model multimodal datasets to verify the effectiveness of our method by comparing it with 15 state-of-the-art methods. Code is available at https://github.com/penghu-cs/RONO.

Existing studies have empirically observed that the resolution of the low-frequency region is easier to enhance than that of the high-frequency one. Although plentiful works have been devoted to alleviating this problem, little understanding is given to explain it. In this paper, we try to give a feasible answer from a machine learning perspective, i.e., the twin fitting problem caused by the long-tailed pixel distribution in natural images. With this explanation, we reformulate image super resolution (SR) as a long-tailed distribution learning problem and solve it by bridging the gaps of the problem between in low- and high-level vision tasks. As a result, we design a long-tailed distribution learning solution, that rebalances the gradients from the pixels in the low- and high-frequency region, by introducing a static and a learnable structure prior. The learned SR model achieves better balance on the fitting of the low- and high-frequency region so that the overall performance is improved. In the experiments, we evaluate the solution on four CNN- and one Transformer-based SR models w.r.t. six datasets and three tasks, and experimental results demonstrate its superiority.

Fair clustering aims to divide data into distinct clusters while preventing sensitive attributes (e.g., gender, race, RNA sequencing technique) from dominating the clustering. Although a number of works have been conducted and achieved huge success recently, most of them are heuristical, and there lacks a unified theory for algorithm design. In this work, we fill this blank by developing a mutual information theory for deep fair clustering and accordingly designing a novel algorithm, dubbed FCMI. In brief, through maximizing and minimizing mutual information, FCMI is designed to achieve four characteristics highly expected by deep fair clustering, i.e., compact, balanced, and fair clusters, as well as informative features. Besides the contributions to theory and algorithm, another contribution of this work is proposing a novel fair clustering metric built upon information theory as well. Unlike existing evaluation metrics, our metric measures the clustering quality and fairness as a whole instead of separate manner. To verify the effectiveness of the proposed FCMI, we conduct experiments on six benchmarks including a single-cell RNA-seq atlas compared with 11 state-of-the-art methods in terms of five metrics. The code could be accessed from https://pengxi.me.

In this paper, we study a novel and widely existing problem in graph matching (GM), namely, Bi-level Noisy Correspondence (BNC), which refers to node-level noisy correspondence (NNC) and edge-level noisy correspondence (ENC). In brief, on the one hand, due to the poor recognizability and viewpoint differences between images, it is inevitable to inaccurately annotate some keypoints with offset and confusion, leading to the mismatch between two associated nodes, i.e., NNC. On the other hand, the noisy node-to-node correspondence will further contaminate the edge-to-edge correspondence, thus leading to ENC. For the BNC challenge, we propose a novel method termed Contrastive Matching with Momentum Distillation. Specifically, the proposed method is with a robust quadratic contrastive loss which enjoys the following merits: i) better exploring the node-to-node and edge-to-edge correlations through a GM customized quadratic contrastive learning paradigm; ii) adaptively penalizing the noisy assignments based on the confidence estimated by the momentum teacher. Extensive experiments on three real-world datasets show the robustness of our model compared with 12 competitive baselines. The code is available at https://github.com/XLearning-SCU/2023-ICCV-COMMON.

In this paper, we study how to achieve two characteristics highly-expected by incomplete multi-view clustering (IMvC). Namely, i) instance commonality refers to that within-cluster instances should share a common pattern, and ii) view versatility refers to that cross-view samples should own view-specific patterns. To this end, we design a novel dual-stream model which employs a dual attention layer and a dual contrastive learning loss to learn view-specific prototypes and model the sample-prototype relationship. When the view is missed, our model performs data recovery using the prototypes in the missing view and the sample-prototype relationship inherited from the observed view. Thanks to our dual-stream model, both cluster- and view-specific information could be captured, and thus the instance commonality and view versatility could be preserved to facilitate IMvC. Extensive experiments demonstrate the superiority of our method on five challenging benchmarks compared with 11 approaches. The code could be accessed from https://pengxi.me.

In recent years, Unsupervised Domain Adaptation (UDA) has emerged as a popular technique for transferring knowledge from a labeled source domain to an unlabeled target domain. However, almost all of the existing approaches implicitly assume that the source domain is correctly labeled, which is expensive or even impossible to satisfy in open-world applications due to ubiquitous imperfect annotations (i.e., noisy labels). In this paper, we reveal that noisy labels interfere with learning from the source domain, thus leading to noisy knowledge being transferred from the source domain to the target domain, termed Dual Noisy Information (DNI). To address this issue, we propose a robust unsupervised domain adaptation framework (ROAD), which prevents the network model from overfitting noisy labels to capture accurate discrimination knowledge for domain adaptation. Specifically, a Robust Adaptive Weighted Learning mechanism (RSWL) is proposed to adaptively assign weights to each sample based on its reliability to enforce the model to focus more on reliable samples and less on unreliable samples, thereby mining robust discrimination knowledge against noisy labels in the source domain. In order to prevent noisy knowledge from misleading domain adaptation, we present a Robust Domain-adapted Prediction Learning mechanism (RDPL) to reduce the weighted decision uncertainty of predictions in the target domain, thus ensuring the accurate knowledge of source domain transfer into the target domain, rather than uncertain knowledge from noise impact. Comprehensive experiments are conducted on three widely-used UDA benchmarks to demonstrate the effectiveness and robustness of our ROAD against noisy labels by comparing it with 13 state-of-the-art methods. Code is available at https://github.com/penghu-cs/ROAD.

Recently, image-text matching has attracted more and more attention from academia and industry, which is fundamental to understanding the latent correspondence across visual and textual modalities. However, most existing methods implicitly assume the training pairs are well-aligned while ignoring the ubiquitous annotation noise, a.k.a noisy correspondence (NC), thereby inevitably leading to a performance drop. Although some methods attempt to address such noise, they still face two challenging problems: excessive memorizing/overfitting and unreliable correction for NC, especially under high noise. To address the two problems, we propose a generalized Cross-modal Robust Complementary Learning framework (CRCL), which benefits from a novel Active Complementary Loss (ACL) and an efficient Self-refining Correspondence Correction (SCC) to improve the robustness of existing methods. Specifically, ACL exploits active and complementary learning losses to reduce the risk of providing erroneous supervision, leading to theoretically and experimentally demonstrated robustness against NC. SCC utilizes multiple self-refining processes with momentum correction to enlarge the receptive field for correcting correspondences, thereby alleviating error accumulation and achieving accurate and stable corrections. We carry out extensive experiments on three image-text benchmarks, i.e., Flickr30K, MS-COCO, and CC152K, to verify the superior robustness of our CRCL against synthetic and real-world noisy correspondences.

In this paper, we study a challenging but less-touched problem in cross-modal retrieval, i.e., partially mismatched pairs (PMPs). Specifically, in real-world scenarios, a huge number of multimedia data (e.g., the Conceptual Captions dataset) are collected from the Internet, and thus it is inevitable to wrongly treat some irrelevant cross-modal pairs as matched. Undoubtedly, such a PMP problem will remarkably degrade the cross-modal retrieval performance. To tackle this problem, we derive a unified theoretical Robust Cross-modal Learning framework (RCL) with an unbiased estimator of the cross-modal retrieval risk, which aims to endow the cross-modal retrieval methods with robustness against PMPs. In detail, our RCL adopts a novel complementary contrastive learning paradigm to address the following two challenges, i.e., the overfitting and underfitting issues. On the one hand, our method only utilizes the negative information which is much less likely false compared with the positive information, thus avoiding the overfitting issue to PMPs. However, these robust strategies could induce underfitting issues, thus making training models more difficult. On the other hand, to address the underfitting issue brought by weak supervision, we present to leverage of all available negative pairs to enhance the supervision contained in the negative information. Moreover, to further improve the performance, we propose to minimize the upper bounds of the risk to pay more attention to hard samples. To verify the effectiveness and robustness of the proposed method, we carry out comprehensive experiments on five widely-used benchmark datasets compared with nine state-of-the-art approaches w.r.t. the image-text and video-text retrieval tasks. The code is available at https://github.com/penghu-cs/RCL.

In this paper, we study how to make unsupervised cross-modal hashing (CMH) benefit from contrastive learning (CL) by overcoming two challenges. To be exact, i) to address the performance degradation issue caused by binary optimization for hashing, we propose a novel momentum optimizer that performs hashing operation learnable in CL, thus making on-the-shelf deep cross-modal hashing possible. In other words, our method does not involve binary-continuous relaxation like most existing methods, thus enjoying better retrieval performance; ii) to alleviate the influence brought by false-negative pairs (FNPs), we propose a Cross-modal Ranking Learning loss (CRL) which utilizes the discrimination from all instead of only the hard negative pairs, where FNP refers to the within-class pairs that were wrongly treated as negative pairs. Thanks to such a global strategy, CRL endows our method with better performance because CRL will not overuse the FNPs while ignoring the true-negative pairs. To the best of our knowledge, the proposed method could be one of the first successful contrastive hashing methods. To demonstrate the effectiveness of the proposed method, we carry out experiments on five widely-used datasets compared with 13 state-of-the-art methods. The code is available at https://github.com/penghu-cs/UCCH.

The success of existing multi-view clustering methods heavily relies on the assumption of view consistency and instance completeness, referred to as the complete information. However, these two assumptions would be inevitably violated in data collection and transmission, thus leading to the so-called Partially View-unaligned Problem (PVP) and Partially Sample-missing Problem (PSP). To overcome such incomplete information challenges, we propose a novel method, termed robuSt mUlti-view clusteRing with incomplEte information (SURE), which solves PVP and PSP under a unified framework. In brief, SURE is a novel contrastive learning paradigm which uses the available pairs as positives and randomly chooses some cross-view samples as negatives. To reduce the influence of the false negatives caused by random sampling, SURE is with a noise-robust contrastive loss that theoretically and empirically mitigates or even eliminates the influence of the false negatives. To the best of our knowledge, this could be the first successful attempt that simultaneously handles PVP and PSP using a unified solution. In addition, this could be one of the first studies on the noisy correspondence problem (i.e., the false negatives) which is a novel paradigm of noisy labels. Extensive experiments demonstrate the effectiveness and efficiency of SURE comparing with 10 state-of-the-art approaches on the multi-view clustering task.

Zero-Shot Sketch-Based Image Retrieval (ZS-SBIR) is a rising theme with broad application prospects. Given the sketch image as a query, the goal of ZS-SBIR is to correctly retrieve the semantically similar images under the zero-shot scenario. The key is to project images from photo and sketch domains into a shared space, where the domain gap and semantic gap are effectively bridged. Most previous studies have approached ZS-SBIR as a classification problem and used classification loss to obtain discriminative features. However, these methods do not explicitly encourage the alignment of features, degrading the retrieval performance. To address this issue, this paper proposes a novel method called Cross-domain Alignment (CA) for ZS-SBIR. Specifically, we present a Large-margin Cross-domain Contrastive (LCC) loss to stimulate intra-class compactness and inter-class separability from both domains, motivated by the relationships of pairwise distances in metric learning. The loss boosts features’ alignment and enjoys more discrimination. Moreover, based on the “embedding stability” phenomenon of the neural network, we elaborate a Cross-batch Semantic Metric (CSM) mechanism for boosting the performance of ZS-SBIR. Extensive experiments demonstrate that the proposed CA achieves encouraging performance on the challenging Sketchy and TU-Berlin benchmarks.

In this paper, we study a challenging problem in image restoration, namely, how to develop an all-in-one method that could recover images from a variety of unknown corruption types and levels. To this end, we propose an All-in-one Image Restoration Network (AirNet) consisting of two neural modules, named Contrastive-Based Degraded Encoder (CBDE) and Degradation-Guided Restoration Network (DGRN). The major advantages of AirNet are two-fold. First, it is an all-in-one solution which could recover various degraded images in one network. Second, AirNet is free from the prior of the corruption types and levels, which just uses the observed corrupted image to perform inference. These two advantages enable AirNet to enjoy better flexibility and higher economy in real world scenarios wherein the priors on the corruptions are hard to know and the degradation will change with space and time. Extensive experimental results show the proposed method outperforms 17 image restoration baselines on four challenging datasets. The code is available at https://github.com/XLearning-SCU/2022-CVPR-AirNet.

In this paper, we study an untouched problem in visible-infrared person re-identification (VI-ReID), namely, Twin Noise Labels (TNL) which refers to as noisy annotation and correspondence. In brief, on the one hand, it is inevitable to annotate some persons with the wrong identity due to the complexity in data collection and annotation, e.g., the poor recognizability in the infrared modality. On the other hand, the wrongly annotated data in a single modality will eventually contaminate the cross-modal correspondence, thus leading to noisy correspondence. To solve the TNL problem, we propose a novel method for robust VI-ReID, termed DuAlly Robust Training (DART). In brief, DART first computes the clean confidence of annotations by resorting to the memorization effect of deep neural networks. Then, the proposed method rectifies the noisy correspondence with the estimated confidence and further divides the data into four groups for further utilizations. Finally, DART employs a novel dually robust loss consisting of a soft identification loss and an adaptive quadruplet loss to achieve robustness on the noisy annotation and noisy correspondence. Extensive experiments on SYSU-MM01 and RegDB datasets verify the effectiveness of our method against the twin noisy labels compared with five state-of-the-art methods. The code could be accessed from https://github.com/XLearning-SCU/2022-CVPR-DART.

Cross-modal retrieval has been a compelling topic in the multimodal community. Recently, to mitigate the high cost of data collection, the co-occurred pairs (e.g., image and text) could be collected from the Internet as a large-scaled cross-modal dataset, e.g., Conceptual Captions. However, it will unavoidably introduce noise (i.e., mismatched pairs) into training data, dubbed noisy correspondence. Unquestionably, such noise will make supervision information unreliable/uncertain and remarkably degrade the performance. Besides, most existing methods focus training on hard negatives, which will amplify the unreliability of noise. To address the issues, we propose a generalized Deep Evidential Cross-modal Learning framework (DECL), which integrates a novel Cross-modal Evidential Learning paradigm (CEL) and a Robust Dynamic Hinge loss (RDH) with positive and negative learning. CEL could capture and learn the uncertainty brought by noise to improve the robustness and reliability of cross-modal retrieval. Specifically, the bidirectional evidence based on cross-modal similarity is first modeled and parameterized into the Dirichlet distribution, which not only provides accurate uncertainty estimation but also imparts resilience to perturbations against noisy correspondence. To address the amplification problem, RDH smoothly increases the hardness of negatives focused on, thus embracing higher robustness against high noise. Extensive experiments are conducted on three image-text benchmark datasets, i.e., Flickr30K, MS-COCO, and Conceptual Captions, to verify the effectiveness and efficiency of the proposed method. The code is available at \urlhttps://github.com/QinYang79/DECL.

Multi-scale architectures have shown effectiveness in a variety of tasks thanks to appealing cross-scale complementarity. However, existing architectures treat different scale features equally without considering the scale-specific characteristics, \textit{i.e.}, the within-scale characteristics are ignored in the architecture design. In this paper, we reveal this missing piece for multi-scale architecture design and accordingly propose a novel Multi-Scale Adaptive Network (MSANet) for single image denoising. Specifically, MSANet simultaneously embraces the within-scale characteristics and the cross-scale complementarity thanks to three novel neural blocks, \textit{i.e.}, adaptive feature block (AFeB), adaptive multi-scale block (AMB), and adaptive fusion block (AFuB). In brief, AFeB is designed to adaptively preserve image details and filter noises, which is highly expected for the features with mixed details and noises. AMB could enlarge the receptive field and aggregate the multi-scale information, which meets the need of contextually informative features. AFuB devotes to adaptively sampling and transferring the features from one scale to another scale, which fuses the multi-scale features with varying characteristics from coarse to fine. Extensive experiments on both three real and six synthetic noisy image datasets show the superiority of MSANet compared with 12 methods. The code could be accessed from https://github.com/XLearning-SCU/2022-NeurIPS-MSANet.

In this article, we study two challenging problems in semisupervised cross-view learning. On the one hand, most existing methods assume that the samples in all views have a pairwise relationship, that is, it is necessary to capture or establish the correspondence of different views at the sample level. Such an assumption is easily isolated even in the semisupervised setting wherein only a few samples have labels that could be used to establish the correspondence. On the other hand, almost all existing multiview methods, including semisupervised ones, usually train a model using a fixed dataset, which cannot handle the data of increasing views. In practice, the view number will increase when new sensors are deployed. To address the above two challenges, we propose a novel method that employs multiple independent semisupervised view-specific networks (ISVNs) to learn representation for multiple views in a view-decoupling fashion. The advantages of our method are two-fold. Thanks to our specifically designed autoencoder and pseudolabel learning paradigm, our method shows an effective way to utilize both the labeled and unlabeled data while relaxing the data assumption of the pairwise relationship, that is, correspondence. Furthermore, with our view decoupling strategy, the proposed ISVNs could be separately trained, thus efficiently handling the data of increasing views without retraining the entire model. To the best of our knowledge, our ISVN could be one of the first attempts to make handling increasing views in the semisupervised setting possible, as well as an effective solution to the noncorresponding problem. To verify the effectiveness and efficiency of our method, we conduct comprehensive experiments by comparing 13 state-of-the-art approaches on four multiview datasets in terms of retrieval and classification.

In many computer vision applications, an object can be represented by multiple different views. Due to the heterogeneous gap triggered by the different views’ inconsistent distributions, it is challenging to exploit these multiview data for cross-view retrieval and classification. Motivated by the fact that both labeled and unlabeled data can enhance the relations among different views, this article proposes a deep cross-view learning framework called deep semisupervised classes- and correlation-collapsed cross-view learning (DSC3L) for cross-view retrieval and classification. Different from the existing methods which focus on the two-view problems, the proposed method learns$U$(generally$U\geq 2$) view-specific deep transformations to gradually project$U$different views into a shared space in which the projection embraces the supervised learning and the unsupervised learning. We propose collapsing the instances of the same class from all views into the same point, with the instances of different classes into distinct points simultaneously. Second, to exploit the abundant unlabeled$U$-wise multiview data, we propose to collapse-correlated data into the same point, with the uncorrelated data into distinct points. Specifically, these two processes are formulated to minimize the two Kullback-Leibler (KL) divergences between the conditional distribution and a desirable one, for each instance. Finally, the two KL divergences are integrated into a joint optimization to learn a discriminative shared space. The experimental results on five widely used public datasets demonstrate the effectiveness of the proposed method.

Image hazing aims to render a hazy image from a given clean one, which could be applied to a variety of practical applications such as gaming, filming, photographic filtering, and image dehazing. To generate plausible haze, we study two less-touched but challenging problems in hazy image rendering, namely, i) how to estimate the transmission map from a single image without auxiliary information, and ii) how to adaptively learn the airlight from exemplars, i.e., unpaired real hazy images. To this end, we propose a neural rendering method for image hazing, dubbed as HazeGEN. To be specific, HazeGEN is a knowledge-driven neural network which estimates the transmission map by leveraging a new prior, i.e., there exists the structure similarity (e.g., contour and luminance) between the transmission map and the input clean image. To adaptively learn the airlight, we build a neural module based on another new prior, i.e., the rendered hazy image and the exemplar are similar in the airlight distribution. To the best of our knowledge, this could be the first attempt to deeply render hazy images in an unsupervised fashion. Compared with existing haze generation methods, HazeGEN renders the hazy images in an unsupervised, learnable, and controllable manner, thus avoiding the labor-intensive efforts in paired data collection and the domain-shift issue in haze generation. Extensive experiments show the promising performance of our method comparing with some baselines in both qualitative and quantitative comparisons. The code is available at https://github.com/XLearning-SCU.

Cross-modal retrieval (CMR) enables flexible retrieval experience across different modalities (e.g., texts versus images), which maximally benefits us from the abundance of multimedia data. Existing deep CMR approaches commonly require a large amount of labeled data for training to achieve high performance. However, it is time-consuming and expensive to annotate the multimedia data manually. Thus, how to transfer valuable knowledge from existing annotated data to new data, especially from the known categories to new categories, becomes attractive for real-world applications. To achieve this end, we propose a deep multimodal transfer learning (DMTL) approach to transfer the knowledge from the previously labeled categories (source domain) to improve the retrieval performance on the unlabeled new categories (target domain). Specifically, we employ a joint learning paradigm to transfer knowledge by assigning a pseudolabel to each target sample. During training, the pseudolabel is iteratively updated and passed through our model in a self-supervised manner. At the same time, to reduce the domain discrepancy of different modalities, we construct multiple modality-specific neural networks to learn a shared semantic space for different modalities by enforcing the compactness of homoinstance samples and the scatters of heteroinstance samples. Our method is remarkably different from most of the existing transfer learning approaches. To be specific, previous works usually assume that the source domain and the target domain have the same label set. In contrast, our method considers a more challenging multimodal learning situation where the label sets of the two domains are different or even disjoint. Experimental studies on four widely used benchmarks validate the effectiveness of the proposed method in multimodal transfer learning and demonstrate its superior performance in CMR compared with 11 state-of-the-art methods.

As Deep Neural Networks (DNNs) usually are overparameterized and have millions of weight parameters, it is challenging to deploy these large DNN models on resource-constrained hardware platforms, e.g., smartphones. Numerous network compression methods such as pruning and quantization are proposed to reduce the model size significantly, of which the key is to find suitable compression allocation (e.g., pruning sparsity and quantization codebook) of each layer. Existing solutions obtain the compression allocation in an iterative/manual fashion while finetuning the compressed model, thus suffering from the efficiency issue. Different from the prior art, we propose a novel One-shot Pruning-Quantization (OPQ) in this paper, which analytically solves the compression allocation with pre-trained weight parameters only. During finetuning, the compression module is fixed and only weight parameters are updated. To our knowledge, OPQ is the first work that reveals pre-trained model is sufficient for solving pruning and quantization simultaneously, without any complex iterative/manual optimization at the finetuning stage. Furthermore, we propose a unified channel-wise quantization method that enforces all channels of each layer to share a common codebook, which leads to low bit-rate allocation without introducing extra overhead brought by traditional channel-wise quantization. Comprehensive experiments on ImageNet with AlexNet/MobileNet-V1/ResNet-50 show that our method improves accuracy and training efficiency while obtains significantly higher compression rates compared to the state-of-the-art.

In this paper, we propose an online clustering method called Contrastive Clustering (CC) which explicitly performs the instance- and cluster-level contrastive learning. To be specific, for a given dataset, the positive and negative instance pairs are constructed through data augmentations and then projected into a feature space. Therein, the instance- and cluster-level contrastive learning are respectively conducted in the row and column space by maximizing the similarities of positive pairs while minimizing those of negative ones. Our key observation is that the rows of the feature matrix could be regarded as soft labels of instances, and accordingly the columns could be further regarded as cluster representations. By simultaneously optimizing the instance- and cluster-level contrastive loss, the model jointly learns representations and cluster assignments in an end-to-end manner. Besides, the proposed method could timely compute the cluster assignment for each individual, even when the data is presented in streams. Extensive experimental results show that CC remarkably outperforms 17 competitive clustering methods on six challenging image benchmarks. In particular, CC achieves an NMI of 0.705 (0.431) on the CIFAR-10 (CIFAR-100) dataset, which is an up to 19% (39%) performance improvement compared with the best baseline. The code is available at https://github.com/XLearning-SCU/2021-AAAI-CC.

Recently, cross-modal retrieval is emerging with the help of deep multimodal learning. However, even for unimodal data, collecting large-scale well-annotated data is expensive and time-consuming, and not to mention the additional challenges from multiple modalities. Although crowd-sourcing annotation, e.g., Amazon’s Mechanical Turk, can be utilized to mitigate the labeling cost, but leading to the unavoidable noise in labels for the non-expert annotating. To tackle the challenge, this paper presents a general Multi-modal Robust Learning framework (MRL) for learning with multimodal noisy labels to mitigate noisy samples and correlate distinct modalities simultaneously. To be specific, we propose a Robust Clustering loss (RC) to make the deep networks focus on clean samples instead of noisy ones. Besides, a simple yet effective multimodal loss function, called Multimodal Contrastive loss (MC), is proposed to maxi-mize the mutual information between different modalities, thus alleviating the interference of noisy samples and cross-modal discrepancy. Extensive experiments are conducted on four widely-used multimodal datasets to demonstrate the effectiveness of the proposed approach by comparing to 14 state-of-the-art methods.

In real-world applications, it is common that only a portion of data is aligned across views due to spatial, temporal, or spatiotemporal asynchronism, thus leading to the so-called Partially View-aligned Problem (PVP). To solve such a less-touched problem without the help of labels, we propose simultaneously learning representation and aligning data using a noise-robust contrastive loss. In brief, for each sample from one view, our method aims to identify its within-category counterparts from other views, and thus the cross-view correspondence could be established. As the contrastive learning needs data pairs as input, we construct positive pairs using the known correspondences and negative pairs using random sampling. To alleviate or even eliminate the influence of the false negatives caused by random sampling, we propose a noise-robust contrastive loss that could adaptively prevent the false negatives from dominating the network optimization. To the best of our knowledge, this could be the first successful attempt of enabling contrastive learning robust to noisy labels. In fact, this work might remarkably enrich the learning paradigm with noisy labels. More specifically, the traditional noisy labels are defined as incorrect annotations for the supervised tasks such as classification. In contrast, this work proposes that the view correspondence might be false, which is remarkably different from the widely-accepted definition of noisy label. Extensive experiments show the promising performance of our method comparing with 10 state-of-the-art multi-view approaches in the clustering and classification tasks. The code will be publicly released at https://pengxi.me.

Non-maximum Suppression (NMS) is an essential post-processing step in modern convolutional neural networks for object detection. Unlike convolutions which are inherently parallel, the de-facto standard for NMS, namely GreedyNMS, cannot be easily parallelized and thus could be the performance bottleneck in convolutional object detection pipelines. MaxpoolNMS is introduced as a parallelizable alternative to GreedyNMS, which in turn enables faster speed than GreedyNMS at comparable accuracy. However, MaxpoolNMS is only capable of replacing the GreedyNMS at the first stage of two-stage detectors like Faster-RCNN. There is a significant drop in accuracy when applying MaxpoolNMS at the final detection stage, due to the fact that MaxpoolNMS fails to approximate GreedyNMS precisely in terms of bounding box selection. In this paper, we propose a general, parallelizable and configurable approach PSRR-MaxpoolNMS, to completely replace GreedyNMS at all stages in all detectors. By introducing a simple Relationship Recovery module and a Pyramid Shifted MaxpoolNMS module, our PSRR-MaxpoolNMS is able to approximate GreedyNMS more precisely than MaxpoolNMS. Comprehensive experiments show that our approach outperforms MaxpoolNMS by a large margin, and it is proven faster than GreedyNMS with comparable accuracy. For the first time, PSRR-MaxpoolNMS provides a fully parallelizable solution for customized hardware design, which can be reused for accelerating NMS everywhere.

Thanks to the low storage cost and high query speed, cross-view hashing (CVH) has been successfully used for similarity search in multimedia retrieval. However, most existing CVH methods use all views to learn a common Hamming space, thus making it difficult to handle the data with increasing views or a large number of views. To overcome these difficulties, we propose a decoupled CVH network (DCHN) approach which consists of a semantic hashing autoencoder module (SHAM) and multiple multiview hashing networks (MHNs). To be specific, SHAM adopts a hashing encoder and decoder to learn a discriminative Hamming space using either a few labels or the number of classes, that is, the so-called flexible inputs. After that, MHN independently projects all samples into the discriminative Hamming space that is treated as an alternative ground truth. In brief, the Hamming space is learned from the semantic space induced from the flexible inputs, which is further used to guide view-specific hashing in an independent fashion. Thanks to such an independent/decoupled paradigm, our method could enjoy high computational efficiency and the capacity of handling the increasing number of views by only using a few labels or the number of classes. For a newly coming view, we only need to add a view-specific network into our model and avoid retraining the entire model using the new and previous views. Extensive experiments are carried out on five widely used multiview databases compared with 15 state-of-the-art approaches. The results show that the proposed independent hashing paradigm is superior to the common joint ones while enjoying high efficiency and the capacity of handling newly coming views.

Cross-modal retrieval aims to retrieve the relevant samples across different modalities, of which the key problem is how to model the correlations among different modalities while narrowing the large heterogeneous gap. In this paper, we propose a Semi-supervised Multimodal Learning Network method (SMLN) which correlates different modalities by capturing the intrinsic structure and discriminative correlation of the multimedia data. To be specific, the labeled and unlabeled data are used to construct a similarity matrix which integrates the cross-modal correlation, discrimination, and intra-modal graph information existing in the multimedia data. What is more important is that we propose a novel optimization approach to optimize our loss within a neural network which involves a spectral decomposition problem derived from a ratio trace criterion. Our optimization enjoys two advantages given below. On the one hand, the proposed approach is not limited to our loss, which could be applied to any case that is a neural network with the ratio trace criterion. On the other hand, the proposed optimization is different from existing ones which alternatively maximize the minor eigenvalues, thus overemphasizing the minor eigenvalues and ignore the dominant ones. In contrast, our method will exactly balance all eigenvalues, thus being more competitive to existing methods. Thanks to our loss and optimization strategy, our method could well preserve the discriminative and instinct information into the common space and embrace the scalability in handling large-scale multimedia data. To verify the effectiveness of the proposed method, extensive experiments are carried out on three widely-used multimodal datasets comparing with 13 state-of-the-art approaches.

In this paper, we study one challenging issue in multi-view data clustering. To be specific, for two data matrices $\mathbf{X}^{(1)}$ and $\mathbf{X}^{(2)}$ corresponding to two views, we do not assume that $\mathbf{X}^{(1)}$ and $\mathbf{X}^{(2)}$ are fully aligned in row-wise. Instead, we assume that only a small portion of the matrices has established the correspondence in advance. Such a partially view-aligned problem (PVP) could lead to the intensive labor of capturing or establishing the aligned multi-view data, which has less been touched so far to the best of our knowledge. To solve this practical and challenging problem, we propose a novel multi-view clustering method termed partially view-aligned clustering (PVC). To be specific, PVC proposes to use a differentiable surrogate of the non-differentiable Hungarian algorithm and recasts it as a pluggable module. As a result, the category-level correspondence of the unaligned data could be established in a latent space learned by a neural network, while learning a common space across different views using the ``aligned'' data. Extensive experimental results show promising results of our method in clustering partially view-aligned data.

Cross-modal retrieval aims to enable flexible retrieval across different modalities. The core of cross-modal retrieval is how to measure the content similarity between different types of data. In this paper, we present a novel cross-modal retrieval method, called Deep Supervised Cross-modal Retrieval (DSCMR). It aims to find a common representation space, in which the samples from different modalities can be compared directly. Specifically, DSCMR minimises the discrimination loss in both the label space and the common representation space to supervise the model learning discriminative features. Furthermore, it simultaneously minimises the modality invariance loss and uses a weight sharing strategy to eliminate the cross-modal discrepancy of multimedia data in the common representation space to learn modality-invariant features. Comprehensive experimental results on four widely-used benchmark datasets demonstrate that the proposed method is effective in cross-modal learning and significantly outperforms the state-of-the-art cross-modal retrieval methods.

Cross-modal hashing, due to its low storage cost and high query speed, has been successfully used for similarity search in multimedia retrieval applications. It projects high-dimensional data into a shared isomorphic Hamming space with similar binary codes for semantically-similar data. In some applications, all modalities may not be obtained or trained simultaneously for some reasons, such as privacy, secret, storage limitation, and computational resource limitation. However, most existing cross-modal hashing methods need all modalities to jointly learn the common Hamming space, thus hindering them from handling these problems. In this paper, we propose a novel approach called Separated Variational Hashing Networks (SVHNs) to overcome the above challenge. Firstly, it adopts a label network (LabNet) to exploit available and nonspecific label annotations to learn a latent common Hamming space by projecting each semantic label into a common binary representation. Then, each modality-specific network can separately map the samples of the corresponding modality into their binary semantic codes learned by LabNet. We achieve it by conducting variational inference to match the aggregated posterior of the hashing code of LabNet with an arbitrary prior distribution. The effectiveness and efficiency of our SVHNs are verified by extensive experiments carried out on four widely-used multimedia databases, in comparison with 11 state-of-the-art approaches.

Cross-modal retrieval takes one type of data as the query to retrieve relevant data of another type. Most of existing cross-modal retrieval approaches were proposed to learn a common subspace in a joint manner, where the data from all modalities have to be involved during the whole training process. For these approaches, the optimal parameters of different modality-specific transformations are dependent on each other and the whole model has to be retrained when handling samples from new modalities. In this paper, we present a novel cross-modal retrieval method, called Scalable Deep Multimodal Learning (SDML). It proposes to predefine a common subspace, in which the between-class variation is maximized while the within-class variation is minimized. Then, it trains m modality-specific networks for m modalities (one network for each modality) to transform the multimodal data into the predefined common subspace to achieve multimodal learning. Unlike many of the existing methods, our method can train different modality-specific networks independently and thus be scalable to the number of modalities. To the best of our knowledge, the proposed SDML could be one of the first works to independently project data of an unfixed number of modalities into a predefined common subspace. Comprehensive experimental results on four widely-used benchmark datasets demonstrate that the proposed method is effective and efficient in multimodal learning and outperforms the state-of-the-art methods in cross-modal retrieval.

In many real-world applications, an object can be described from multiple views or styles, leading to the emerging multi-view analysis. To eliminate the complicated (usually highly nonlinear) view discrepancy for favorable cross-view recognition and retrieval, we propose a Multi-view Linear Discriminant Analysis Network (MvLDAN) by seeking a nonlinear discriminant and view-invariant representation shared among multiple views. Unlike existing multi-view methods which directly learn a common space to reduce the view gap, our MvLDAN employs multiple feedforward neural networks (one for each view) and a novel eigenvalue-based multi-view objective function to encapsulate as much discriminative variance as possible into all the available common feature dimensions. With the proposed objective function, the MvLDAN could produce representations possessing: 1) low variance within the same class regardless of view discrepancy, 2) high variance between different classes regardless of view discrepancy, and 3) high covariance between any two views. In brief, in the learned multi-view space, the obtained deep features can be projected into a latent common space in which the samples from the same class are as close to each other as possible (even though they are from different views), and the samples from different classes are as far from each other as possible (even though they are from the same view). The effectiveness of the proposed method is verified by extensive experiments carried out on five databases, in comparison with the 19 state-of-the-art approaches.