Is this you? Claim this profile to correct it, add a bio and choose the work people see first.
Claim this profileWorks272 from public data
- R³Net: Recurrent Residual Refinement Network for Saliency Detection522
A novel recurrent residual refinement network (R^3Net) equipped with residual refinement blocks (RRBs) to more accurately detect salient regions of an input image that outperforms competitors in all the benchmark datasets.
- Leveraging the Invariant Side of Generative Zero-Shot Learning347
A novel method is proposed which can directly generate the unseen features from random noises which are conditioned by the semantic descriptions, and is named leveraging invariant side GAN (LisGAN), which can outperform state-of-the-art methods with significant improvements.
- Maximum Density Divergence for Domain Adaptation342
This paper proposes a new domain adaptation method named adversarial tight match (ATM) which enjoys the benefits of both adversarial training and metric learning and proposes a novel distance loss, named maximum density divergence (MDD), to quantify the distribution divergence.
- A Comprehensive Survey on Source-Free Domain Adaptation308
A comprehensive survey of recent advances in Source-Free Domain Adaptation (SFDA) is provided and several components of each method are modularized to more clearly illustrate their relationships and mechanisms in light of the composite properties of each method.
- 299
- Locality Preserving Joint Transfer for Domain Adaptation274
This work proposes a novel approach to jointly exploit feature adaptation with distribution matching and sample adaptation with landmark selection, suitable for both homogeneous- and heterogeneous-domain adaptations by learning domain-specific projections.
- MMALFM235
A proposed multi-modal aspect-aware topic model (MATM) is applied on text reviews and item images to model users’ preferences and items’ features from different aspects, and also estimates the aspect importance of a user toward an item to alleviate the data sparsity problem and gain good interpretability for recommendation.
- 216
- Cross-Domain Gradient Discrepancy Minimization for Unsupervised Domain Adaptation212
A cross-domain gradient discrepancy minimization (CGDM) method which explicitly minimizes the discrepancy of gradients generated by source samples and target samples so it can be used as a good supervision to improve the accuracy of target samples.
- 186
- 174
- 172
- 169
- DA-GCN: A Domain-aware Attentive Graph Convolution Network for Shared-account Cross-domain Sequential Recommendation162
This work proposes a novel graph-based solution, namely DA-GCN, to address the challenges of shared-account Cross-domain Sequential Recommendation, and devise a domain-aware graph convolution network to learn user-specific node representations.
- Exploring Auxiliary Context: Discrete Semantic Transfer Hashing for Scalable Image Retrieval158
This paper proposes a novel hashing approach, dubbed as discrete semantic transfer hashing (DSTH), to directly augment the semantics of discrete image hash codes by exploring auxiliary contextual modalities and guarantees direct semantic transfer and avoid information loss.
- Cross-Modal Retrieval: A Systematic Review of Methods and Future Directions152
A comprehensive review of cross-modal retrieval’s evolution, spanning from shallow statistical analysis techniques to vision-language pretraining (VLP) models, and delves deeply into the principles and architectures underpinning existing cross-modal retrieval methods.
- Dynamic Multi-View Hashing for Online Image Retrieval130
This paper proposes dynamic multi-view hashing (DMVH), which can adaptively augment hash codes according to dynamic changes of image, and demonstrates superior performance of DWVH over several state-of-the-art hashing methods.
- Discrete Multimodal Hashing With Canonical Views for Robust Mobile Landmark Search128
A new augmented Lagrangian multiplier (ALM) based optimization method to directly solve the discrete binary codes and can desirably avoid accumulated quantization errors in conventional optimization method which simply adopts two-step ``relaxing+rounding'' framework.
- Multi-Behavior Recommendation with Cascading Graph Convolution Networks127
A novel multi-behavior recommendation model with cascading graph convolution networks (named MB-CGCN), in which the embeddings learned from one behavior are used as the input features for the next behavior's embedding learning after a feature transformation operation.
- 127
- 125
- 125
- Explicit Interaction Model towards Text Classification121
A novel framework, EXplicit interAction Model (dubbed as EXAM), equipped with the interaction mechanism to incorporate word-level matching signals into the text classification task is designed.
- 120
- Online Cross-Modal Hashing for Web Image Retrieval120
This paper proposes Online Cross-modal Hashing (OCMH) which can effectively address the above two problems by learning the shared latent codes (SLC) of hash codes and dynamic transfer matrix, and demonstrates the effectiveness and efficiency of OCMH for online cross- modal web image retrieval.
- 102
- Unsupervised Topic Hypergraph Hashing for Efficient Mobile Image Retrieval102
A novel unsupervised hashing scheme, called topic hypergraph hashing (THH), is proposed, to address the semantic shortage of hashing codes by exploiting auxiliary texts around images and can achieve superior performance compared with several state-of theart methods.
- 100
- Mutual-Enhanced Incongruity Learning Network for Multi-Modal Sarcasm Detection99
A mutual enhancement module is introduced to take advantage of the underlying consistency between the two modules to boost the performance of the Mutual-enhanced Incongruity Learning Network for multi-modal sarcasm detection, named MILNet.
- Content-Based Visual Landmark Search via Multimodal Hypergraph Learning93
This work proposes multimodal hypergraph (MMHG) to characterize the complex associations between landmark images and designs a novel content-based visual landmark search system based on MMHG to facilitate effective search.
- 87
- 87
- Multi-view feature selection via Nonnegative Structured Graph Learning83
A unified learning framework is developed, which directly learns the structured graph from the raw features by imposing a rank constraint, and simultaneously performs adaptive feature selection with exploiting the complementarity of multi-view features.
- 80
- 79
- An Attribute-Aware Attentive GCN Model for Attribute Missing in Recommendation78
An attribute-aware attentive graph convolution network capable of incorporating associate attributes to strengthen the user and item representation learning, and thus naturally solve the attribute missing problem is presented.
- Domain Adaptive Semantic Segmentation without Source Data76
This work proposes an effective framework for domain adaptive semantic segmentation without source data, which assumes that the model is pre-trained on the source domain, and then adapting to the target domain without accessing source data anymore.
- 75
- Cycle-consistent Conditional Adversarial Transfer Networks75
This paper proposes a novel approach named cycle-consistent conditional adversarial transfer networks (3CATN), which takes care of the domain alignment by leveraging adversarial training and condition the adversarial networks with the cross-covariance of learned features and classifier predictions to capture the multimodal structures of data distributions.
- Adaptive Collaborative Similarity Learning for Unsupervised Multi-view Feature Selection73
This paper proposes an adaptive collaborative similarity learning (ACSL) for multi-view feature selection to dynamically learn the collaborative similarity structure, and further integrate it with the ultimate feature selection into a unified framework.
- Discrete Optimal Graph Clustering71
A structured similarity graph that is theoretically optimal for clustering performance is adaptively learned with a guidance of reasonable rank constraints and an iterative optimization strategy guaranteed with convergence is developed to directly solve the clustering results.
- Time Interval-Enhanced Graph Neural Network for Shared-Account Cross-Domain Sequential Recommendation68
A new graph-based solution, namely, time interval-enhanced domain-aware graph convolutional network (TiDA-GCN), to address the challenges of shared-account cross-domain sequential recommendation and fully account for users’ domain-specific preferences on items.
- 66
- 66
- Flexible Multi-modal Hashing for Scalable Multimedia Retrieval61
A novel Flexible Multi-modal Hashing method that learns multiple modality-specific hash codes and multi- modality collaborative hash codes simultaneously within a single model to address the problem of binarize queries when only one or part of modalities are provided.
- 60
- 59
- 58
- Unsupervised soft-label feature selection55
An effective Unsupervised Soft-label Feature Selection (USFS) model is proposed, which performs soft-label learning and simultaneously guides the unsupervised feature selection process with the learned soft-labels.
- 55
- 54
- 54
- 51
- 51
- 50
- Dual-Level Semantic Transfer Deep Hashing for Efficient Social Image Retrieval50
This paper designs a complementary dual-level semantic transfer mechanism to efficiently discover the potential semantics of tags and seamlessly transfer them into binary hash codes and achieves superior performance compared with state-of-the-art hashing methods.
- SCOTCH and SODA: A Transformer Video Shadow Detection Framework49
This work introduces the shadow deformation attention trajectory (SODA), a new type of video self-attention module, specially designed to handle the large shadow deformations in videos, and presents a new shadow contrastive learning mechanism (SCOTCH) which aims at guiding the network to learn a unified shadow representation from massive positive shadow pairs across different videos.
- 49
- Unsupervised multi-graph cross-modal hashing for large-scale multimedia retrieval49
MGCMH is unsupervised method which integrates multi-graph learning and hash function learning into a joint framework, to learn unified hash space for all modalities, and an alternating learning algorithm is proposed to jointly optimize the modality weights, hash codes and functions.
- 48
- 47
- Efficient discrete latent semantic hashing for scalable cross-modal retrieval47
Extensive experiments show that, DLSH outperforms several state-of-the-art cross-modal hashing methods and directly correlates the explicit semantic labels with binary codes, which increases efficiency and reduces the quantization loss in hash optimization.
- Exploring Consistent Preferences45
This paper proposes a novel discrete hashing with pair-exemplar (DHPE) to support scalable and efficient large-scale CBVLS and designs a novel optimization method based on augmented Lagrange multiplier to explicitly deal with the discrete constraint, the bit-uncorrelated constraint and balance constraint.
- Efficient Query-based Black-box Attack against Cross-modal Hashing Retrieval44
A multi-modal knockoffs-driven adversarial generation is proposed to achieve efficient adversarial example generation and experiments show that EQB2A achieves superior attacking performance under the black-box setting.
- 42
- 42
- Unsupervised Deep Cross-modal Hashing with Virtual Label Regression42
This paper proposes a novel unified learning framework to jointly perform deep hash function training, virtual label learning and regression and incorporates thevirtual label learning into the objective functions and simultaneously regress the learned virtual labels to the hash codes.
- mCLOUD: A Multiview Visual Feature Extraction Mechanism for Ground-Based Cloud Image Categorization42
The experiments on a challenging cloud dataset termed the six-class Huazhong University of Science and Technology (HUST) cloud demonstrate that mCLOUD consistently outperforms other visual feature extraction algorithms in categorizing clouds based on ground-based images.
- 41
- Source-Free Active Domain Adaptation via Energy-Based Locality Preserving Transfer41
This work proposes a Locality Preserving Transfer (LPT) framework which preserves and utilizes locality structures on target data to achieve adaptation without source data and shows that this method is comparable or superior to current state-of-the-art active domain adaptation methods even without access to source data.
- 41
- Challenging tough samples in unsupervised domain adaptation41
A novel approach, named challenging tough sample networks (CTSN), to practice the three ideas for domain adaptation and tame tough samples, which is able to handle both easy and tough samples.
- 40
- 39
- Learning compact visual representation with canonical views for robust mobile landmark search38
A Canonical View based Compact Visual Representation (2CVR) is proposed to handle mobile Landmark Search problems via novel three-stage learning which preserves visual relations of images measured with canonical views and removes noises.
- 36
- MMPOI: A Multi-Modal Content-Aware Framework for POI Recommendations34
This paper proposes a novel multi-modal content-aware framework for POI recommendation (MMPOI), which leverages pre-trained models for inter-modal conversion and employs a unified pre-trained model to extract modal-specific features from each modality, effectively bridging the semantic gap between different modalities.
- 34
- Prototype-guided Knowledge Transfer for Federated Unsupervised Cross-modal Hashing34
PT-FUCH protects local private data by exploring unified global prototypes for different clients, without relying on any semantic annotations, and designs an adaptive cross-modal knowledge distillation to transfer valuable semantic knowledge from modal-specific global models to local prototype learning processes, reducing the risk of overfitting.
- Weighting scheme for image retrieval based on bag‐of‐visual‐words34
This study presents a novel combining method which captures query-specific weights for visual-words in query image according to their discriminative information and demonstrates that the proposed weighting scheme can obtain higher retrieval performance than other weighting schemes.
- 33
- Adversarial cross-modal retrieval based on dictionary learning33
A novel framework termed Adversarial Cross-Modal Retrieval Based on Dictionary Learning Algorithm (DLA-CMR), where dictionary learning serves as feature reconstructor to reconstruct discriminative features, while adversarial learning mines the statistical characteristics for each modality.
- BadCM: Invisible Backdoor Attack Against Cross-Modal Learning32
This paper introduces a novel bilateral backdoor to fill in the missing pieces of the puzzle in the cross-modal backdoor and proposes a generalized invisible backdoor framework against cross-modal learning (BadCM), the first invisible backdoor method deliberately designed for diverse cross-modal attacks within one unified framework.
- Effective Comparative Prototype Hashing for Unsupervised Domain Adaptation32
This paper establishes a domain-shared unit hypersphere space through prototype contrastive learning and then obtain the Hamming hypersphere space via mapping from the shared hypersphere and achieves a cohesive synergy between learning uniformly distributed and category conflict-averse feature representations, eliminating domain discrepancies, and facilitating hash code learning.
- 32
- 32
- Cross-Modal Self-Taught Hashing for large-scale image retrieval32
Cross-Modal Self-Taught Hashing (CMSTH) is proposed for large-scale cross-modal and unimodal image retrieval and can effectively capture the semantic correlation from unlabeled training data.
- Motion Prediction via Joint Dependency Modeling in Phase Space31
This paper introduces a novel convolutional neural model to effectively leverage explicit prior knowledge of motion anatomy, and simultaneously capture both spatial and temporal information of joint trajectory dynamics, and proposes a global optimization module that learns the implicit relationships between individual joint features.
- SOIL: Contrastive Second-Order Interest Learning for Multimodal Recommendation30
This work proposes a Second-Order Interest Learning (SOIL) framework to retrieve second-order interest from unrecorded suboptimal items, and optimize the model by implementing contrastive learning modules for user and item representations at both the user-item and item-item levels.
- Shifting More Attention to Breast Lesion Segmentation in Ultrasound Videos30
A novel frequency and localization feature aggregation network (FLA-Net) that learns temporal features from the frequency domain and predicts additional lesion location positions to assist with breast lesion segmentation is proposed.
- 30
- Unsupervised multi-view feature extraction with dynamic graph learning30
A unified learning framework is devised which simultaneously performs dynamic graph learning and the feature extraction and demonstrates the superior performance of the proposed approach, compared with state-of-the-art techniques.
- 29
- Weighted locality collaborative representation based on sparse subspace29
A face recognition approach named Weighted Locality Collaborative Representation Classifier based on sparse subspace (WLCRC), which firstly learns a subset of the original training data to build a much correlated dictionary, and then combines linear regression techniques together with weighted collaborative representation techniques to optimize the linear reconstruction of unlabeled data.
- DA-DAN: A Dual Adversarial Domain Adaption Network for Unsupervised Non-overlapping Cross-domain Recommendation28
To let DA-DAN be free of auxiliary information, it learns users’ preferences by only exploring their sequential patterns, and proposes an improved self-attention layer to model them, and resorts to a dual generative adversarial network with a multi-target adversarial loss.
- 28
- 28
- Topic Hypergraph Hashing for Mobile Image Retrieval28
A novel unsupervised hashing scheme, THH, is proposed, which can achieve superior performance compared with several state-of-the-art techniques, and it is more suitable for MIR.
- 27
- Identity-Guided Collaborative Learning for Cloth-Changing Person Reidentification27
A novel identity-guided collaborative learning scheme (IGCL) for cloth-changing person ReID is proposed, where the human semantic is effectively utilized and the identity is unchangeable to guide collaborative learning.
- Efficient weakly-supervised discrete hashing for large-scale social image retrieval27
A unified weakly-supervised hash learning framework that could effectively enrich the semantics of image hash codes with the freely obtained user-provided social tags and simultaneously remove their involved adverse noises.
- Scribble Hides Class: Promoting Scribble-Based Weakly-Supervised Semantic Segmentation with Its Class Label26
This study proposes a class-driven scribble promotion network, which utilizes both scribble annotations and pseudo-labels informed by image-level classes and global semantics for supervision, and introduces a distance entropy loss for uncertainty reduction.
- 26
- Efficient Multi-modal Hashing with Online Query Adaption for Multimedia Retrieval25
A Supervised Multi-modal Hashing with Online Query-adaption method designed to adaptively preserve the multi-modal features into hash codes by exploiting their complementarity and a parameter-free online hashing module to adaptively capture query variations at the online retrieval stage is proposed.
- Exploring the Cross-Domain Action Recognition Problem by Deep Feature Learning and Cross-Domain Learning25
Deep feature learning method has much better generalization ability than the hand-crafted feature, such as improved dense trajectory if there are enough labeled samples in the training dataset to be used to fine-tune the network, and both unsupervised cross-domain learning method and supervised cross- domain learning method can improve the performance.
- 24
- 24
- Cross-domain Recommendation via Dual Adversarial Adaptation24
The proposed approach builds a dual adversarial adaptation (DAA) framework to adversarially train the target model together with a pre-trained source model to alleviate imbalanced sample distribution and thus challenge the data scarcity issue in the target domain.
- Learning explicitly transferable representations for domain adaptation24
This work proposes to adapt classifiers rather than features to fill in the distribution gaps between domains by some additional transferable representations which are explicitly learned from the original features while keeping theOriginal features unchanged.
- Incomplete Cross-Modal Retrieval with Deep Correlation Transfer23
This article proposes an Incomplete Cross-Modal Retrieval with Deep Correlation Transfer (ICMR-DCT) method that can robustly model incomplete multi-modal data and dynamically capture the adjacency semantic correlation for cross-modal retrieval.
- I read, I saw, I tell23
A novel FGVC approach called Texts Assisted Fine-Grained Visual Classification (TA-FGVC) is proposed, which reads from texts to gain attention, sees the images with the gained attention and then tells the subtle differences.
- Cross-media retrieval with collective deep semantic learning23
This paper proposes an effective retrieval method, dubbed as Cross-media Retrieval with Collective Deep Semantic Learning (CR-CDSL), to solve the problem of cross-media retrieval, where two complementary deep neural networks are first learned to collectively project image and text samples into a joint semantic representation.
- Place-centric Visual Urban Perception with Deep Multi-instance Regression23
Results with comparisons showed that the proposed method can clearly outperform the alternative perception methods and more importantly, is capable of generating region-level safety scores to facilitate interpretations of the perception process.
- 21
- 21
- Pixel Exclusion: Uncertainty-aware Boundary Discovery for Active Cross-Domain Semantic Segmentation21
This paper proposes a novel Uncertainty-aware Boundary Discovery (UBD) strategy that selects the uncertain pixels in the boundary areas that contains rich contextual information and outperforms previous methods with both pixel-level and image-level label acquisition protocols.
- JDMAN: Joint Discriminative and Mutual Adaptation Networks for Cross-Domain Facial Expression Recognition21
This work designs a mutual information minimization module for domain-level adaptation, which narrows the domain shift by simultaneously distilling the domain-invariant components and eliminating the untransferable ones lying in different domains, and proposes a semantic metric learning module for category-level adaptations, which can close the semantic discrepancy during discriminative intra-domain representation learning and transferable inter-domain knowledge discovery.
- Invisible Black-Box Backdoor Attack against Deep Cross-Modal Hashing Retrieval20
A flexible trigger generator to generate the attacker’s specified triggers, which learns the sample semantics of the non-poisoned modality to bridge the cross-modal attack gap, and an input-aware injection network, which embeds the generated triggers into benign samples in the form of sample-specific stealth and realizes cross-modal semantic interaction between triggers and poisoned samples.
- Disentangled Cascaded Graph Convolution Networks for Multi-Behavior Recommendation20
The Disentangled Cascaded Graph Convolutional Network (Disen-CGCN), a novel multi-behavior recommendation model that employs disentangled representation techniques to effectively separate factors within user and item representations, ensuring their independence, is proposed.
- 20
- 20
- 20
- 20
- MCRPL: A Pretrain, Prompt, and Fine-tune Paradigm for Non-overlapping Many-to-one Cross-domain Recommendation19
This work focuses on NMCR, a non-overlapping MCR scenario, and devise MCRPL, a cross-domain recommendation method that improves the recommendations in the sparse target domain by leveraging the information from other rich domains.
- Fusion-Supervised Deep Cross-Modal Hashing19
A novel Fusion-supervised Deep Cross-modal Hashing (FDCH) approach that learns unified binary codes through a fusion hash network with paired samples as input, which effectively enhances the modeling of the correlation of heterogeneous multi- modal data.
- DDPO: Direct Dual Propensity Optimization for Post-Click Conversion Rate Estimation18
A Direct Dual Propensity Optimization (DDPO) framework to optimize the model directly in impression space with both clicked and unclicked samples and to strengthen the causal relationship, two causal transfer modules for the conversion rate prediction model with the attention mechanism are designed.
- Task-adaptive Asymmetric Deep Cross-modal Hashing18
A Task- Adaptive Asymmetric Deep Cross-modal Hashing (TA-ADCMH) method that can learn task-adaptive hash functions for two sub-retrieval tasks via simultaneous modality representation and asymmetric hash learning and is proved on two standard datasets from many aspects.
- Robust optimal graph clustering18
This paper proposes a novel clustering framework where a robust graph is learned with noise removal, and simultaneously, with desirable clustering structure to solve the formulated unified learning framework to achieve better prediction accuracy.
- Label consistent locally linear embedding based cross-modal hashing18
LCLCH preserves the non-linear manifold structure of different modality data by Locally Linear Embedding, and transforms heterogeneous data into a latent common semantic space to reduce the semantic gap and support cross-modal retrieval tasks.
- A Stochastic Attribute Grammar for Robust Cross-View Human Tracking18
A learning method to train this attribute grammar model from weakly supervised training data is developed and an alternative inference method to employ both topdown and bottom-up computations to search the optimal solution is developed.
- A New Dew and Frost Detection Sensor Based on Computer Vision18
A novel method for indirectly detecting dew and frost based on computer vision is proposed, which is simple, inexpensive, and only requires images of several glass substrates near the underlying surface.
- Effective naive Bayes nearest neighbor based image classification on GPU18
This paper explores accelerating strategies from perspectives of both algorithm design and software development and integrates space decomposition capability of Product quantization (PQ) and parallel accelerating capability of underlying computational platform, Graphics processing unit (GPU).
- 17
- 17
- Two-pronged Strategy17
This paper proposes a simple and efficient Lightweight Augmented Graph Network Hashing (LAGNH) method with a two-pronged strategy, which designs a cross-modal attention module based on the auxiliary semantic information to adaptively mitigate the adverse effects in the deep image features.
- Multi-level Attention-based Domain Disentanglement for BCDR16
This article proposes a Multi-level Attention-based Domain Disentanglement framework dubbed MADD for BCDR, which explicitly leverages the attention mechanism to construct personalized preference with both domain-invariant and domain-specific features obtained by disentangling raw user embeddings.
- 16
- 16
- Federated Semantic Learning for Privacy-preserving Cross-domain Recommendation15
This work introduces federated semantic learning and devise FFMSR, a Fast Fourier Transform-based filter and a gating mechanism are developed to alleviate the impact of irrelevant semantic information in the local model.
- Joint feature selection and graph regularization for modality-dependent cross-modal retrieval15
The key idea of JFGM is learning modality-dependent subspaces for different sub-retrieval tasks while simultaneously preserving the semantic consistency of multi-modal data.
- Privacy-Enhanced Prototype-Based Federated Cross-Modal Hashing for Cross-Modal Retrieval14
In PEPFCH, local and global prototypes are integrated in order to effectively capture the distinctive traits of individual clients, while also harnessing the collective intelligence of the entire federated learning system to facilitate personalized federated learning.
- 14
- 14
- Locality-Constrained Transfer Coding for Heterogeneous Domain Adaptation14
This work proposes to deploy locality-constrained transfer coding (LCTC) to simultaneously alleviate the feature discrepancy and mitigate the distribution divergence, and aims at tackling heterogeneous domain adaptation (HDA).
- Improved information gain feature selection method for Chinese text classification based on word embedding14
An improved feature selection method is proposed which uses word embedding to calculate the most similar words to the current dictionary selected by IG algorithm and expand the dictionary with these words under certain regulations.
- 13
- Task-Dependent and Query-Dependent Subspace Learning for Cross-Modal Retrieval13
This work develops a unified cross-modal learning framework, where task-specific and category-specific subspaces can be learned simultaneously via an efficient iterative optimization and demonstrates the superior performance of the proposed approach compared with several state-of-the-art techniques.
- Diff-VPS: Video Polyp Segmentation via a Multi-task Diffusion Network with Adversarial Temporal Reasoning12
This paper incorporates multi-task supervision into diffusion models to promote the discrimination of diffusion models on pixel-by-pixel segmentation and equip TRM with a generative adversarial self-supervised strategy to produce more realistic frames and thus capture better dynamic cues.
- 12
- 12
- A Novel and Effective Model to Predict Skip Metastasis in Papillary Thyroid Carcinoma Based on a Support Vector Machine12
The model can accurately distinguish the skip metastasis in PTC using a simple and affordable method, which may have potential for daily clinical application in the future.
- Learning Transferrable and Interpretable Representations for Domain Generalization12
This paper aims to learn a domain transformation space via a domain transformer network (DTN) which explicitly mines the relationship among multiple domains and constructs transferable feature representations for down-stream tasks by interpreting each feature as a semantically weighted combination of multiple domain-specific features.
- Image segmentation using modified SLIC and Nyström based spectral clustering12
An efficient two-stage image segmentation method is proposed which takes advantage of modified SLIC segmentation and Nystrom based spectral clustering and the memory requirement and the computational complexity are significantly reduced.
- Adaptive Collaborative Soft Label Learning for Unsupervised Multi-View Feature Selection11
An Adaptive Collaborative Soft Label Learning (ACSLL) model for unsupervised multi-view feature selection is proposed, which learns the pseudo soft labels from each view feature by a simple and efficient method and fuse them with an adaptive weighting strategy into a joint soft label matrix.
- Noise-Robust Continual Test-Time Domain Adaptation11
This paper addresses the noise-robustness problem in continual TTA by offering three effective recipes to mitigate it, and proposes a soft ensemble negative learning mechanism to guide the model optimization using ensemble complementary pseudo labels.
- 11
- 11
- 10
- Unified Modality Separation: A Vision-Language Framework for Unsupervised Domain Adaptation10
This work proposes a unified modality separation framework that accommodates both modality-specific and modality-invariant components and designs a modality discrepancy metric to categorize samples into modality-invariant, modality-specific, and uncertain ones.
- Lite-Mind: Towards Efficient and Robust Brain Representation Learning10
Lite-Mind is proposed, a lightweight, efficient, and robust brain representation learning paradigm based on Discrete Fourier Transform (DFT), which efficiently aligns fMRI voxels to fine-grained information of CLIP.
- 10
- Adaptive and flexible $$\ell _1$$-norm graph embedding for unsupervised feature selection9
This work proposes an efficient alternative updating algorithm with an iterative reweighted strategy, together with the necessary convergence and complexity analyses, to solve the challenging problem of unsupervised feature selection.
- 9
- Webly Supervised Image Hashing with Lightweight Semantic Transfer Network9
A Webly Supervised Image Hashing with a well-designed lightweight network that enhances the semantics of unsupervised image hashing with the weak supervision from freely available web images, and simultaneously avoids involving over-abundant parameters in the deep network architecture.
- 8
- 8
- Discriminative Visual Similarity Search with Semantically Cycle-consistent Hashing Networks8
This article proposes a novel deep tripartite semantically interactive hashing framework, dubbed Semantically Cycle-consistent Hashing Networks (SCHNs), for discriminative hash code learning, and establishes the cyclic principle of deep semantic-preserving hashing by adaptive semantic parsing across different spaces in a single-modal visual similarity search.
- 8
- 7
- 7
- 7
- Semantic-enhanced Co-attention Prompt Learning for Non-overlapping Cross-domain Recommendation7
The prompt learning technique for Many-to-one Non-overlapping Non-overlapping Cross-domain Sequential Recommendation (MNCSR) is introduced and a Text-enhanced Co-attention Prompt Learning Paradigm (TCPLP) is proposed, which captures semantic meanings by representing items through text rather than IDs, leveraging natural language universality to facilitate cross-domain knowledge transfer.
- 7
- 7
- 7
- 7
- 7
- An efficient multi-feature SVM solver for complex event detection7
A self-adaptive multi-feature learning framework with efficient Support Vector Machine (SVM) solver for complex event detection in this paper that is able to utilize multiple features reasonably with an adaptively weighted linear combination manner, which is simple yet effective.
- Learning Robust Graph Hashing for Efficient Similarity Search7
A Robust Graph Hashing (RGH) is proposed, which automatically learns robust graph based on self-representation of samples to alleviate the noises and seamlessly integrates graph construction and hashing learning into a unified learning framework.
- 7
- Learning Together Securely: Prototype-Based Federated Multi-Modal Hashing for Safe and Efficient Multi-Modal Retrieval6
Prototype-based Federated Multi-modal Hashing is proposed, an innovative framework that seamlessly integrates federated learning with multi-modal hashing techniques, and achieves fine-grained fusion of heterogeneous multi-modal data, enhancing retrieval accuracy while ensuring data privacy through prototype-based communication, thereby reducing communication costs and mitigating risks of data leakage.
- 6
- 6
- ARTS: A General and Efficient Multi-Task Self-Prompt Framework for Explainable Sequential Recommendation6
A self-prompt generator that transfers the user’s global behavior features into the continuous prompt, achieving efficient information sharing among multi-task modules and significantly improves the model’s time and space efficiency is proposed.
- 6
- 6
- Instance-level Adversarial Source-free Domain Adaptive Person Re-identification6
An Instance-level Adversarial Mutual Teaching (IAMT) framework is introduced, which utilizes adversarial views to tackle the challenges mentioned above and reveals the superiority of the IAMT over state-of-the-art methods for DA person ReID tasks, while preserving the privacy of the source data.
- 6
- 6
- 5
- 5
- PatAug: Augmentation of Augmentation for Test-Time Adaptation5
Patch Augmentation is introduced, a pixel-level perturbation framework that optimizes the benefits of informative augmentations and mitigates negative transformation impacts, and adjusts patches with a cross-modal similarity alignment loss and learnable patching weights to better fit the target distributions.
- 5
- 5
- 5
- Accelerating Content-Based Image Retrieval via GPU-Adaptive Index Structure5
A novel graphics processing unit (GPU) adaptive index structure, termed as plane semantic ball (PSB), is proposed to simultaneously reduce the work of retrieval process and exploit the parallel acceleration of underlying hardware.
- 4
- 4
- 4
- 4
- 4
- 4
- 4
- A method of frost observation based on intensity changing regularity simulation and texture analysis4
The method can get high detection accuracy in the different kinds of continuous changing environment and the characteristic of frost surface texture is also well described by texture analysis based on texture descriptor.
- 4
- 3
- Flip is Better than Noise: Unbiased Interest Generation for Multimedia Recommendation3
The approach aims to generate valid latent user interests while non-invasively preserving the original interest graph structure and significantly improves NDCG@20 by 7.3% on the TikTok dataset and boosts Recall@20 by 4.2% on the Sports dataset.
- 3
- 3
- Efficient inter-image relation graph neural network hashing for scalable image retrieval3
This paper proposes a simple but effective Inter-image Relation Graph Neural Network Hashing (IRGNNH) method, which significantly outperforms the state-of-the-art unsupervised deep hashing methods on both retrieval accuracy and efficiency.
- 3
- Token-Level Prompt Mixture With Parameter-Free Routing for Federated Domain Generalization2
TRIP is proposed, a Token-level pRompt mIxture with Parameter-free routing framework for FedDG that treats prompts as multiple experts, and assigns individual tokens within an image to distinct experts, facilitating the capture of fine-grained visual patterns.
- EffiPOI: A Product Quantization Framework Based on Knowledge Distillation for Efficient POI Recommendations2
A hybrid knowledge distillation strategy that transfers both response-aware and feature-aware knowledge from the teacher model to the student model is developed, to alleviate the performance degradation due to quantization.
- 2
- 2
- 2
- 2
- 2
- 2
- 2
- 2
- Task-Adversarial Adaptation for Multi-modal Recommendation2
A Task-Adversarial Adaptation (TAA) framework, which is able to align data distributions and reduce resource consumption at the same time, and formulate the proposed approach as a plug-and-play module to accelerate the model training and improve the performance of mainstream multi-modal multi-task recommendation systems.
- 2
- 2
- 2
- Multi-Task Multi-modal Semantic Hashing for Web Image Retrieval with Limited Supervision2
Multi-Task Multi-modal Semantic Hashing (MTMSH) is proposed to index large scale social image data collection with limited supervision and improves search accuracy via improving more semantic information from two aspects.
- Learning Robust Multi-Label Hashing for Efficient Image Retrieval2
A label hypergraph is constructed to effectively capture high-order semantic correlations of images and preserved into hashing codes with hypergraph consistency and direct label-hashing correlation and an efficient algorithm based on Alternate Direction Method of Multipliers is developed to calculate the optimal hashing codes.
- 1
- Timing Analysis of the TSN Time-aware Shaper Based on Compositional Performance Analysis1
The optimized CPA analysis of the worst-case end-to-end delay of class-based TAS streams in vehicular networks greatly reduces conservative error and can be used to validate the appropriateness of the class-based gate control parameter configurations.
- Noise-Robust Generative Hashing for Cross-Modal Retrieval1
Noise-Robust Generative Hashing (NRGH), a unified framework designed to accommodate various forms of noise inherent in cross-modal retrieval, is proposed and significantly exceeds state-of-the-art baselines in various noisy scenarios, delivering superior robustness and accuracy.
- 1
- Learning Semantic Facial Descriptors for Accurate Face Animation1
This work introduces the semantic facial descriptors in learnable disentangled vector space to address the issue of model-based methods' limitations in high-fidelity identity and the challenges faced by model-free methods in accurate motion transfer.
- 1
- 1
- ANAGL: A Noise-Resistant and Anti-Sparse Graph Learning for Micro-Video Recommendation1
This work proposes a noise-resistant and anti-sparse graph learning framework for micro-video recommendation that employs a multi-view reconstruction approach and utilize cross-view self-supervised learning to gain insights into user and micro-video features.
- 1
- 1
- 1
- 1
- 1
- 1
- –
- –
- –
- –
- HGDNet: De-Noised Review-Based Rating Prediction Using Hierarchical Gating and Discriminative Networks–
A novel Hierarchical Gating and Discriminative model for rating prediction that utilizes personalized end-to-end differential thresholds to select reviews in a relatively “hard” manner, thereby minimizing the impact of noisy reviews while facilitating model training.
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- Sentiment Classification for Chinese Product Reviews Based on Semantic Relevance of Phrase–
A new sentiment analysis algorithm is utilized to analyze sentiment of Chinese product reviews and results show that when dealing with massive data, the algorithm is better than traditional algorithms on both accuracy and learning time.
- Reconstructing In-Silico Gene Regulatory Network Based on Reproducing Kernel Hilbert Space–
Comparative experimental results reveal that the HSIC method has stronger capability of structural identification and is feasible for construct complex gene regulatory network.
- Numerical simulation of fire risk of large-scale supermarket–
Taking a supermarket as example, the temperature and height of smoke were studied according to whether the sprinkler was functioning using CFAST software to show that occupants could evacuate safely from the supermarket.
- –
- –
- –
Publication data from OpenAlex, with missing venues and authors filled in from Crossref; citation counts are the higher of OpenAlex and Semantic Scholar, last synced 2026-10-14. One-sentence summaries under some papers are written by Semantic Scholar’s model. Citation counts may be lower than on Google Scholar, which indexes more sources.
Report an error
Wrong papers, two people merged into one, or a profile that should not be here? Tell us and we will fix or hide it.