Is this you? Claim this profile to correct it, add a bio and choose the work people see first.
Claim this profileWorks172 from public data
- Deep Domain Generalization via Conditional Invariant Adversarial Networks848
This work proposes an end-to-end conditional invariant deep domain generalization approach by leveraging deep neural networks for domain-invariant representation learning and proves the effectiveness of the proposed method.
- The Sixth Visual Object Tracking VOT2018 Challenge Results804
The Visual Object Tracking challenge VOT2018 is the sixth annual tracker benchmarking activity organized by the VOT initiative; results of over eighty trackers are presented; many are state-of-the-art trackers published at major computer vision conferences or in journals in the recent years.
- A Twofold Siamese Network for Real-Time Object Tracking672
The proposed SA-Siam outperforms all other real-time trackers by a large margin on OTB-2013/50/100 benchmarks and proposes a channel attention mechanism for the semantic branch.
- A Multi-Organ Nucleus Segmentation Challenge558
Several of the top techniques compared favorably to an individual human annotator and can be used with confidence for nuclear morphometrics as well as heavy data augmentation in the MoNuSeg 2018 challenge.
- 496
- Quantization Networks432
This paper provides a simple and uniform way for weights and activations quantization by formulating it as a differentiable non-linear function that will shed new lights on the interpretation of neural network quantization.
- Gaussian Temporal Awareness Networks for Action Localization369
Gaussian Temporal Awareness Networks (GTAN) is presented --- a new architecture that novelly integrates the exploitation of temporal structure into an one-stage action localization framework and achieves superior results when comparing to state-of-the-art approaches.
- Prompt Distribution Learning367
This work presents prompt distribution learning for effectively adapting a pre-trained vision-language model to address downstream recognition tasks and employs a Gaussian distribution to model them effectively and derive a surrogate loss for efficient training.
- Exploring Object Relation in Mean Teacher for Cross-Domain Detection355
This work presents Mean Teacher with Object Relations (MTOR) that novelly remolds Mean Teacher under the backbone of Faster R-CNN by integrating the object relations into the measure of consistency cost between teacher and student modules.
- Domain Generalization via Conditional Invariant Representations312
This paper proposes to learn a feature representation which has domain-invariant class conditional distributions P(h(X)|Y), and proposes a conditional invariant representation which can be guaranteed if the class prior P(Y) does not change across training and test domains.
- Graph Edge Convolutional Neural Networks for Skeleton-Based Action Recognition205
The dynamics of human limbs for skeleton-based action recognition is investigated, and an edge in a graph of a human skeleton is represented by integrating its spatial neighboring edges and its temporal neighboring edges for achieving the consistency of movements in an action.
- Learning Spatio-Temporal Representation With Local and Global Diffusion198
A novel framework to boost the spatio-temporal representation learning by Local and Global Diffusion (LGD) is presented, which construct a novel neural network architecture that learns the local and global representations in parallel.
- DisPFL: Towards Communication-Efficient Personalized Federated Learning via Decentralized Sparse Training186
This work proposes a novel personalized federated learning framework in a decentralized (peer-to-peer) communication protocol named Dis-PFL, which employs personalized sparse masks to customize sparse local models on the edge to save the communication and computation cost.
- Camera Lens Super-Resolution170
This paper investigates SR from the perspective of camera lenses, named as CameraSR, which aims to alleviate the intrinsic tradeoff between resolution (R) and field-of-view (V) in realistic imaging systems and quantitatively analyzes the performance of commonly-used synthetic degradation models.
- Bayesian video search reranking161
Two new methods are proposed in this paper to measure the ranking distance based on the disagreement in terms of pair-wise orders, which represents the disagreement between the objective ranking list and the initial text-based.
- 158
- 132
- Sparse transfer learning for interactive video search reranking119
This article proposes a novel dimension reduction tool, termed sparse transfer learning (STL), to effectively and efficiently encode user's labeling information and conducts extensive experiments on the TRECVID 2005, 2006 and 2007 benchmark datasets and compared STL with popular dimension reduction algorithms.
- DigestPath: A benchmark dataset with challenge review for the pathological detection and segmentation of digestive-system117
Two well-annotated benchmark datasets and organized challenges for the digestive-system pathological cell detection and tissue segmentation are released, in conjunction with the International Conference on Medical Image Computing and Computer-Assisted Intervention (MICCAI).
- Deep Boosting for Image Denoising110
A novel deep boosting framework (DBF) for denoising, which integrates several convolutional networks in a feed-forward fashion, and introduces the concept of dense connection that overcomes the vanishing of gradients during training.
- QAOA-in-QAOA: Solving Large-Scale MaxCut Problems on Small Quantum Machines103
This work revisits the MaxCut problem via the divide-and-conquer heuristic and proposes QAOA-in-QAOA, a method to solve arbitrary large-scale MaxCut problems using small quantum machines that can be seamlessly embedded into other advanced strategies to enhance the capability of QAOAs in large- scale combinatorial optimization problems.
- 103
- 103
- On Efficient Training of Large-Scale Deep Learning Models102
This survey presents a comprehensive review of the general techniques for training acceleration and provides a detailed analysis and discussion of future works on the development of general acceleration techniques, which could inspire us to re-think and design novel efficient paradigms.
- Mocycle-GAN101
A new Motion-guided Cycle GAN, dubbed as Mocycle-GAN, that novelly integrates motion estimation into unpaired video translator that capitalizes on three types of constrains: adversarial constraint discriminating between synthetic and real frame, cycle consistency encouraging an inverse translation on both frame and motion, and motion translation validating the transfer of motion between consecutive frames.
- From Yes-Men to Truth-Tellers: Addressing Sycophancy in Large Language Models with Pinpoint Tuning78
This work proposes a novel supervised pinpoint tuning (SPT), where the region-of-interest modules are tuned for a given objective, and demonstrates that SPT significantly mitigates the sycophancy issue of LLMs (even better than SFT).
- Spatio-Temporal Inception Graph Convolutional Networks for Skeleton-Based Action Recognition78
This work designs a simple and highly modularized graph convolutional network architecture for skeleton-based action recognition by repeating a building block that aggregates multi-granularity information from both the spatial and temporal paths.
- Continuous Dropout.78
The proposed continuous dropout is considerably closer to the activation characteristics of neurons in the human brain than traditional binary dropout and has the property of avoiding the co-adaptation of feature detectors, which suggests that it can extract more independent feature detectors for model averaging in the test stage.
- Photo Quality Assessment with DCNN that Understands Image Well77
A deep convolutional neural network is adopted that “understands” images well and is proved to be effective in many computer vision problems and it does not need human efforts in the design of features.
- 69
- 66
- Interpreting and Improving Large Language Models in Arithmetic Calculation63
This work serves as a preliminary exploration into the arithmetic calculation abilities inherent in LLMs, laying a solid foundation to reveal more intricate mathematical tasks.
- 62
- Towards a Better Match in Siamese Network Based Visual Object Tracker59
This paper proposes two simple yet effective mechanisms, namely angle estimation and spatial masking, to address issues of Siamese network based trackers, and produces a tracker, named Siam-BM, which is the best-performing realtime tracker to date.
- On Efficient Training of Large-Scale Deep Learning Models: A Literature Review58
This survey presents a comprehensive review to understand the general mechanisms within each component and their joint interaction of training deep learning models, and presents this comprehensive taxonomy.
- Toward Real-world Single Image Deraining: A New Benchmark and Beyond56
The experimental results show the difference of representative methods in image restoration performance and model complexity, validate the significance of the proposed datasets for model generalization, and provide useful insights on the superiority of learning from diverse domains and shed lights on the future research on real-world SID.
- 54
- A Style and Semantic Memory Mechanism for Domain Generalization49
It is verified that it is critical for the network to be informative on what domain features are invariant and shared among in-stances, so that the network sharpens its understanding and improves its semantic discriminative ability.
- Local Convolutional Neural Networks for Person Re-Identification49
For the task of person re-identification, the local convolutional neural networks (Local CNN) can outperform state-of-the-art methods consistently on three large-scale benchmarks, including Market-1501, CUHK03, and DukeMTMC-ReID.
- On Better Exploring and Exploiting Task Relationships in Multitask Learning: Joint Model and Feature Learning47
This paper proposes a new strategy to measure the relatedness that jointly learns shared parameters and shared feature representations and conducts various experiments to verify the superiority of the proposed joint model and feature MTL method.
- 46
- Multi-task proximal support vector machine46
This paper proposes a novel multi-task learning model based on the proximal support vector machine that achieves better performance compared with other popular multi- task learning models and develops a method to optimize the learning procedure of MTPSVM.
- Multi-level photo quality assessment with multi-view features46
A set of compact rule-based features through careful analyses on photographic rules and aesthetic attributes are designed and proposed using Deep Convolutional Neural Network descriptor, which encodes the photo content thoroughly, to implicitly describe the photo quality.
- Transform-Invariant Convolutional Neural Networks for Image Classification and Search44
This paper proposes randomly transforming feature maps of CNNs during the training stage to prevent complex dependencies of specific rotation, scale, and translation levels of training images in CNN models and shows that random transformation provides significant improvements ofCNNs on many benchmark tasks, including small-scaleimage recognition, large-scale image recognition, and image retrieval.
- Revisiting Knowledge Distillation: An Inheritance and Exploration Framework39
This paper proposes a novel inheritance and exploration knowledge distillation framework (IE-KD), in which a student model is split into two parts - inheritance and Exploration, which can be a general technique to improve the student network to achieve SOTA performance.
- Towards Theoretical Understandings of Self-Consuming Generative Models37
Borders are derived on the total variation (TV) distance between the synthetic data distributions produced by future models and the original real data distribution under various mixed training scenarios for diffusion models with a one-hidden-layer neural network score function.
- Concentrated Local Part Discovery With Fine-Grained Part Representation for Person Re-Identification36
- Multi-task model and feature joint learning36
A theoretical bound is presented which directly demonstrates that the proposed multi-task learning method can successfully model the relatedness via joint common parameter- and common feature-learning.
- A Multiscale-Grid-Based Stacked Bidirectional GRU Neural Network Model for Predicting Traffic Speeds of Urban Expressways35
The experimental results show that the stacked BiGRU network with the multiscale-grid model enables to capture the spatial-temporal characteristics of traffic speed efficiently and outperforms benchmark models in the prediction of the traffic speed.
- 34
- Moderately Distributional Exploration for Domain Generalization33
MODE performs distribution exploration in an uncertainty $\textit{subset}$ that shares the same semantic factors with the training domains and can endow models with provable generalization performance on unknown target domains.
- Video Retrieval with Similarity-Preserving Deep Temporal Hashing33
This article proposes an end-to-end video retrieval framework called the Similarity-Preserving Deep Temporal Hashing (SPDTH) network, which unifies video temporal modeling and learning to hash into one step to allow for maximum retention of information.
- 33
- 31
- Transferrable Contrastive Learning for Visual Domain Adaptation30
This work presents a particular paradigm of self-supervised learning tailored for domain adaptation, i.e., Transferrable Contrastive Learning (TCL), which links the SSL and the desired cross-domain transferability congruently, and finds contrastive learning intrinsically a suitable candidate fordomain adaptation.
- Deep Domain Adaptation Hashing with Adversarial Learning30
A novel Deep Domain Adaptation Hashing with Adversarial learning (DeDAHA) architecture is presented, which mainly consists of three components: a deep convolutional neural networks for learning basic image/frame representation followed by an adversary stream on one hand to optimize the domain discriminator, and on the other, to interact with each domain-specific hashing stream for encoding image representation to hash codes.
- Constrained Metric Learning Via Distance Gap Maximization30
This paper poses robust and tractable metric learning under pairwise constraints that are expressed as similarity judgements between data pairs, and finds that the constrained metric learning approach has widespread applicability without being limited to particular backgrounds.
- Multi-source Multi-level Attention Networks for Visual Question Answering29
This work proposes multi-source multi-level attention networks for visual question answering that can benefit both spatial inferences by visual attention on context-aware region representation and reasoning by semantic attention on concepts as well as external knowledge.
- FedImpro: Measuring and Improving Client Update in Federated Learning28
FedImpro decouples the model into high-level and low-level components, and trains the high-level portion on reconstructed feature distributions that enhances the generalization contribution and reduces the dissimilarity of gradients in FL.
- 28
- Flickr group recommendation using rich social media information28
This paper proposes a group recommendation scheme utilizing users' trust neighbors and similar neighbors' tastes and tries to find users' similar neighbors using tag information, not only from users' photos but also from their favorite photos and the common friend information.
- 26
- Query aware visual similarity propagation for image search reranking26
This paper proposes to estimate a query aware image similarity by incorporating the global visual similarity, local visual similarity and visual word co-occurrence into an iterative propagation framework.
- A Theoretical Perspective: How to Prevent Model Collapse in Self-consuming Training Loops23
This paper introduces the intriguing notion of recursive stability and presents the first theoretical generalization analysis, revealing how both model architecture and the proportion between real and synthetic data influence the success of STLs and extends this analysis to transformers in in-context learning.
- 23
- Classification and Representation Joint Learning via Deep Networks21
A deep learning framework that considers both class label information and local spatial distribution information between training samples is proposed, and a two-channel network with shared weights is used to measure the local distribution.
- 21
- Learning to judge image search results21
This paper proposes a novel method to automatically identify the best search result list from a number of candidates, and proposes a preference learning model to quantitatively study the best image search result identification problem.
- Dual-Path Distillation: A Unified Framework to Improve Black-Box Attacks20
A novel framework, dual-path distillation, is proposed that utilizes the feedback knowledge not only to craft adversarial examples but also to alter the searching directions to achieve efficient attacks.
- 20
- 19
- 19
- 19
- MSRA atT TRECVID 2008: High-Level Feature Extraction and Automatic Search.19
This paper describes the MSRA experiments for TRECVID 2008, which representatively investigated the benefit of global and local low-level features by a variety of learning-based methods, including supervised and semi-supervised learning algorithms.
- Click-through-based Subspace Learning for Image Search18
It is demonstrated that the above two fundamental challenges can be mitigated by jointly exploring the subspace learning and the use of click-through data and the former aims to create a latent subspace with the ability in comparing information from the original incomparable views.
- Domain-Class Correlation Decomposition for Generalizable Person Re-Identification17
This model outperforms the state-of-the-art methods on the large-scale domain generalization Re-ID benchmark and can learn a domain-invariant representation that can capture more class-related features.
- Learning to Localize Actions from Moments17
This paper introduces a new design of transfer learning type to learn action localization for a large set of action categories, but only on action moments from the categories of interest and temporal annotations of untrimmed videos from a small set of Action Herald Networks (AherNet).
- 16
- 15
- 15
- Visual Reranking with Local Learning Consistency15
This paper proposes to use local learning regularizer to model the multiple-wise consistency, by formulating the consistent score estimation over a local area into a learning problem.
- Robust Training of Federated Models with Extremely Label Deficiency14
This work proposes a novel twin-model paradigm, called Twin-sight, designed to enhance mutual guidance by providing insights from different perspectives of labeled and unlabeled data, and introduces a neighbourhood-preserving constraint that encourages the preservation of the neighbourhood relationship among data features extracted by both models.
- A heuristic multi-objective task scheduling framework for container-based clouds via actor-critic reinforcement learning14
A heuristics multi-objective task scheduling framework based on reinforcement learning (AC-CCTS) that not only solves the problems of single objective and local convergence in traditional task scheduling methods but also reduces the cost of experiential learning with reinforcement learning methods.
- 14
- Learning Relative Aesthetic Quality with a Pairwise Approach14
A pairwise-based ranking framework that takes image pairs as input to address the challenge of aesthetic quality assessment and shows clear advantages over traditional binary classification-based approach.
- Leveraging Submodule Linearity Enhances Task Arithmetic Performance in LLMs13
This work argues that linearity already exists within the model's submodules, and proposes an innovative model merging strategy that independently merges these submodules to derive a closed-form solution for optimal merging weights grounded in the linear properties of these submodules.
- Out-of-Distribution Detection with Virtual Outlier Smoothing13
A novel method, Virtual Outlier Smoothing (VOSo), is proposed, which constructs auxiliary outliers using ID samples, thereby eliminating the need to search for OOD samples.
- 13
- 13
- 13
- 13
- 13
- 13
- 13
- 13
- Good Questions Help Zero-Shot Image Reasoning12
Question-Driven Visual Exploration (QVix) is introduced, a novel prompting strategy that enhances the exploratory capabilities of LVLMs in zero-shot reasoning tasks and significantly outperforms existing methods, highlighting its effectiveness in bridging the gap between complex visual data and LVL Ms' exploratory abilities.
- KCNN: Kernel-wise Quantization to Remarkably Decrease Multiplications in Convolutional Neural Network12
This paper quantizes the floating-point weights in each kernel separately to multiple bit planes to remarkably decrease multiplications and proposes dual normalization to solve the pathological curvature problem during fine-tuning.
- 12
- 12
- 11
- 11
- An Active Service Recommendation Model for Multi-Source Remote Sensing Information Using Fusion of Attention and Multi-Perspective10
This paper proposes a recommendation algorithm using fusion of attention and multi-perspective (MRS_AMRA) for massive multi-source remote sensing resources by combining streaming pushing technology and develops a prototype system to verify the effectiveness and timeliness of MRS_ASRM.
- Bi-calibration Networks for Weakly-Supervised Video Representation Learning10
A new design of mutual calibration between query and text to achieve more reliable visual-textual supervision for video representation learning is introduced and Bi-Calibration Networks (BCN) is presented that novelly couples two calibrations to learn the correction from text to query and vice versa.
- Joints kinetic and relational features for action recognition10
A set of comprehensive features, termed joints kinetic and relational features (JKRF), for action recognition, and the experimental results show that JKRF consistently outperform state-of-the-art action recognition methods.
- 10
- 10
- 10
- Enhancing Target-unspecific Tasks through a Features Matrix9
This work proposes a novel Features Matrix (FM) approach designed to enhance these models on target-unspecific tasks, which captures the semantics of diverse inputs from a deep and fine perspective, preserving essential general knowledge, which mitigates the risk of overfitting.
- 9
- Estimation Method of Intersection Signal Cycle Based on Empirical Data9
The signal cycle plays an important role in the design of timing optimization and evaluation for a signal intersection and in some cities, there is a lack of unified management of this cycle.
- Learning multi-granularity features from multi-granularity regions for person re-identification9
A novel person re-identification network to learn discriminative features across multiple granularities from body regions which are also multi-grained is proposed, which is the state-of-the-art in recent years.
- Query difficulty estimation via relevance prediction for image retrieval9
A novel query difficulty estimation approach which automatically estimates the average precision of the image search results through an adaptive pseudo positive image selection method and an efficient voting scheme to estimate images' relevance probability.
- 3D Creation at Your Fingertips: From Text or Image to 3D Assets8
An automatic 3D creation system, which can create realistic 3D assets solely from a text or image prompt without requiring any specialized 3D modeling skills, and a coarse-to-fine framework to achieve this goal.
- Animating Your Life8
An unified unpaired video-to-video translator is utilized to explore both appearance structure and temporal continuity in video synthesis, which can transform real-world video into cartoon or ink-wash animation in real-time.
- Multi-modal and multi-scale photo collection summarization8
A multi-modal and multi-scale photo collection summarization method by leveraging multi- modal features, including time, location and high-level semantic features, and a novel key photo ranking and selection algorithm that takes the importance of both events and photos into consideration.
- Towards Lightweight Black-Box Attacks against Deep Neural Networks7
It is argued that black-box attacks can pose practical attacks in this extremely restrictive scenario where only several test samples are available and the proposed Error TransFormer (ETF) transforms the approximation error in the parameter space into a perturbation in the feature space and alleviates the error by disturbing features.
- Eigenvector-Based Distance Metric Learning for Image Classification and Retrieval7
This work proposes a novel distance metric learning approach by introducing base vectors explicitly learned from training data that can reflect the intrinsic information of the input features and can extend to any kernel space without changing the objective function.
- 7
- Patch Reordering: A NovelWay to Achieve Rotation and Translation Invariance in Convolutional Neural Networks7
It is shown that the patch ranking module improves the performance of the CNN on many benchmark tasks, including MNIST digit recognition, large-scale image recognition, and image retrieval.
- 7
- 7
- Magnetic tracker calibration for an augmented reality system for therapy7
The results demonstrate that the methods used to compensate distortions produced by metallic objects which are moving during the measurement procedure, but fixed with a receiver are capable of significantly reducing quasi-static distortions.
- Advancing Prompt Learning through an External Layer6
This work proposes a paradigm called EnPrompt with a novel External Layer (EnLa), which proposes a textual external layer and learnable visual embeddings for adapting VLMs to downstream tasks and outperforms the existing prompt learning method.
- Interpretable Composition Attribution Enhancement for Visio-linguistic Compositional Understanding6
This work investigates the cause for deficient behavior in composition attribution issue and shows such issue is mitigated via a novel framework called CAE (Composition Attribution Enhancement), which incorporates various interpretable attribution methods to encourage the model to pay greater attention to composition words denoting relationships and attributes within the text.
- 6
- Semantic-Aware Mixup for Domain Generalization6
This work proposes to perform a semantic-aware mixup (SAM) for domain generalization, where whether to perform mixup depends on the semantic and domain information, and proves the effectiveness of SAM using image classification tasks on several DG benchmarks.
- Discriminative codebook learning for Web image search6
A novel discriminative codebook learning method is proposed by introducing the subspace learning in codebook construction and leveraging its power to find a contextual local descriptor subspace to capture the discrim inative information.
- Towards Generalizable Detector for Generated Image5
A novel perspective on understanding and improving generated image detection is proposed, inspired by the human cognitive process, and it is demonstrated that models trained merely over natural images bring guaranteed generalization ability under mild assumptions.
- Semi-supervised Coefficient-Based Distance Metric Learning5
Considering the situation that labeled samples are insufficient in some cases, the proposed distance metric learning method is extended into a semi-supervised learning framework and an optimization algorithm is proposed to accelerate training of the methods.
- 5
- 4
- 4
- CRAR: Accelerating Stereo Matching with Cascaded Residual Regression and Adaptive Refinement4
This article integrates refinement methods to create an unified algorithm to utilize parallelism for running devices to further accelerate the inference of stereo matching algorithms through the improvement of network structure.
- 4
- HRP: High-Rank Preheating for Superior LoRA Initialization3
High-Rank Preheating (HRP) is proposed, which first trains LoRA with a higher preheating rank for a few steps, then uses the main singular vectors of the derived $BA^\top$ as initialization for the main fine-tuning process, and it is proved that HRP makes LoRA achieve better fine-tuned results than random initialization in expectation.
- Epistemic Uncertainty for Generated Image Detection3
A novel framework for AI-generated image detection through epistemic uncertainty, aiming to address critical security concerns in the era of generative models by utilizing advanced large-scale vision models pre-trained on extensive natural images to estimate the epistemic uncertainty of images and flag those with high uncertainty as generated.
- Boosting Fair Classifier Generalization through Adaptive Priority Reweighing3
A novel adaptive reweighing method that prioritizes samples closer to the decision boundary and assigns a higher weight to improve the generalizability of fair classifiers to eliminate the impact of the distribution shifts between training and test data on model generalizability.
- 3
- 3
- 3
- Sequence-to-Sequence Learning via Shared Latent Representation3
This paper proposes a star-like framework for general and flexible sequence-to-sequence learning, where different types of media contents could be encoded to and decoded from a shared latent representation (SLR) (the central node).
- A Small Scale Multi-Column Network for Aesthetic Classification Based on Multiple Attributes3
This paper designs an extremely small network with only 38 K parameters for better training and proposes a multi-column network architecture to embed prior information into the deep learning model, showing that prior information indeed guides the network to learn better.
- Visual Re-ranking Through Greedy Selection and Rank Fusion3
This paper first conducts label de-noising to filter false-positive images, then a simple greedy graph-based re-ranking algorithm is proposed to derive the resulting list and using the rank fusion, the results from different graphs are combined to form a better result.
- A New SVM Iterative Algorithm in Large Training Set3
A new SVM iterative algorithm is proposed, aiming at the problem that the speeds of learning and classifying are slow in large training set, and the scales of training data and the support vector set are effectively compressed.
- Bridging the Language Gap: Uncovering and Aligning Shared Circuits for Multi-Hop Reasoning in Multilingual LLMs2
This work introduces a mechanistic interpretability framework that traces the causal pathways of multi-hop knowledge reasoning, and reveals a core, non-obvious finding: cross-lingual inconsistencies do not stem from a knowledge deficit.
- 2
- 2
- Layer-Wise Training to Create Efficient Convolutional Neural Networks2
This paper proposes an algorithm to reduce computational cost, which is often solved by sparsification and matrix decomposition methods, and replaces the complex convolutional layers in large networks with more efficient modules and keep their outputs in each-layer consistent.
- 2
- 2
- 2
- 2
- Cross-Modal Masked Compositional Concept Modeling for Enhancing Visio-Linguistic Compositionality1
This work proposes MACCO (MAsked Compositional Concept MOdeling), a framework that masks compositional concepts in one modality and reconstructs them conditioned on the full contextual information from the other, enabling the model to capture and align cross-modal compositional structures more effectively.
- 1
- Structured Cooperative Learning with Graphical Model Priors1
Structured Cooperative Learning (SCooL), in which a cooperation graph across devices is generated by a graphical model prior to automatically coordinate mutual learning between devices, is proposed and compared with existing decentralized learning methods on an extensive set of benchmarks.
- 1
- 1
- 1
- 1
- 1
- –
- –
- –
- Drawback of Enforcing Equivariance and its Compensation via the Lens of Expressive Power–
Surprisingly, it is shown that the enlarged neural architectures have reduced hypothesis space dimensionality, implying even better generalizability.
- An Effective Levelling Paradigm for Unlabeled Scenarios–
The proposed LePa, designed as a plug-in module, dynamically constrains and coordinates multiple objective functions, thereby improving the generalization of these baseline methods.
- –
- –
- Unsupervised person re-identification based on removal of camera bias and dynamic updating of the memory bank–
This work combines the removal of the camera bias and the dynamic updating of the memory model to better solve problems in the unsupervised person reidentification field.
- Partially Annotated Gastric Pathological Image Classification–
A cascaded convolutional neural network framework is proposed to classify partially annotated pathological images to detect cancer regions, which are re-identified by a patch-wise classification network.
- –
- –
- –
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.