AAAI 2025 Papers — Page 29
AAAI Conference on Artificial Intelligence · 3028 papers
Towards Scalable and Deep Graph Neural Networks via Noise Masking
Yuxuan Liang (Peking University), Bin Cui (Wuhan University)
Graph Neural NetworkGraph
🎯 What it does: A pluggable RMask module is proposed to suppress over-smoothing and enhance deep propagation effects in simplified graph neural networks through noise masking and random walks.
Towards Ship License Plate Recognition in the Wild: A Large Benchmark and Strong Baseline
Baolong Liu (Zhejiang Gongshang University), Jianfeng Dong (Zhejiang Gongshang University)
RecognitionTransformerContrastive LearningImageBenchmark
🎯 What it does: A large ship license plate recognition dataset SLP34K has been constructed, and a recognition baseline based on self-supervised pre-training and semantic enhancement has been proposed.
Towards Trustable SHAP Scores
Olivier Létoffé (University of Toulouse), Joao Marques-Silva (Institute for Catalan Research and Advanced Studies)
Explainability and InterpretabilityTabular
🎯 What it does: This paper proposes several new feature functions (υ_s, υ_a, υ_c, υ_n, etc.) by modifying the feature function of SHAP, and defines the corresponding SHAP scores based on this to eliminate unreasonable feature importance results that occur in existing SHAP calculations.
Towards Trustworthy Knowledge Graph Reasoning: An Uncertainty Aware Perspective
Bo Ni (Vanderbilt University), Tyler Derr (Vanderbilt University)
Knowledge DistillationHyperparameter SearchGraph Neural NetworkLarge Language ModelTextGraph
🎯 What it does: This paper proposes the UAG (Uncertainty Aware Knowledge Graph Reasoning) framework, which utilizes multi-step reasoning combining KG and LLM, and calibrates the error rates of each component through confidence prediction and Learn-Then-Test, forming a theoretically guaranteed set of answers.
Towards Unbiased Information Extraction and Adaptation in Cross-Domain Recommendation
Yibo Wang (Nanjing University), Lijun Zhang (Nanjing University)
Recommendation SystemAuto EncoderTabular
🎯 What it does: A multi-objective cross-domain recommendation method UIEA is proposed, which can achieve better recommendation results in data-sparse domains by extracting unbiased information and adapting it.
Towards Unifying Evaluation of Counterfactual Explanations: Leveraging Large Language Models for Human-Centric Assessments
Marharyta Domnich (Institute of Computer Science, University of Tartu), Raul Vicente (Institute of Computer Science, University of Tartu)
Explainability and InterpretabilityTransformerLarge Language ModelSupervised Fine-TuningPrompt EngineeringText
🎯 What it does: Collected 30 adversarial explanations, invited 206 participants to score them across 8 dimensions, and utilized a large language model (LLM) to predict human evaluations.
Towards Universal Rainy Image Restoration: Benchmark and Baseline
Hujie Yan (California Institute of Technology)
RestorationObject DetectionConvolutional Neural NetworkImageBenchmark
🎯 What it does: This paper proposes a Universal Rainy Image Restoration task (URIR), constructs a high-quality URIR-8K dataset, and designs a baseline model based on multi-scale Mamba to handle four types of rain damage patterns with a single model.
Towards Verifiable Text Generation with Generative Agent
Bin Ji (National University of Defense Technology), See-Kiong Ng (National University of Singapore)
GenerationRetrievalTransformerLarge Language ModelTextBenchmarkRetrieval-Augmented GenerationChain-of-Thought
🎯 What it does: A retrieval and reflection-based generative agent R²-MGA is proposed, which can dynamically select the best matching contextual examples in text generation and citation tasks and refine answers through multiple rounds.
Toy-GS: Assembling Local Gaussians for Precisely Rendering Large-Scale Free Camera Trajectories
Xiaohan Zhang (South China University of Technology), Qi Liu (South China University of Technology)
GenerationOptimizationComputational EfficiencyGaussian SplattingPoint Cloud
🎯 What it does: The Toy-GS method is proposed, which utilizes adaptive spatial partitioning, Patchmatch and PPAC, as well as local-global fusion techniques to achieve high-quality rendering of large-scale free camera trajectories, significantly reducing GPU memory usage.
TPCH: Tensor-interacted Projection and Cooperative Hashing for Multi-view Clustering
Zhongwen Wang (Nanjing University of Science and Technology), Zhenwen Ren (Sichuan University)
OptimizationAuto EncoderImage
🎯 What it does: The TPCH (Tensor-Interacted Projection and Cooperative Hashing) framework is proposed, which stacks multi-view projection matrices and hash codes into tensors, introducing high-order cooperation in the projection and Hamming space and enhancing the tensor nuclear norm to improve the density and separability of binary representations for large-scale multi-view clustering.
TRACI: A Data-centric Approach for Multi-Domain Generalization on Graphs
Yusheng Zhao (Peking University), Ming Zhang (Peking University)
ClassificationDomain AdaptationData-Centric LearningGraph Neural NetworkContrastive LearningGraphBenchmark
🎯 What it does: This work studies domain generalization of multi-source graph data and proposes the TRACI method to enhance the generalization performance of GNNs without a target graph.
Track the Answer: Extending TextVQA from Image to Video with Spatio-Temporal Clues
Yan Zhang (Institute of Information Engineering, Chinese Academy of Sciences), Can Ma (Institute of Information Engineering, Chinese Academy of Sciences)
RecognitionGenerationTransformerLarge Language ModelVision Language ModelVideoText
🎯 What it does: For the task of Video Text Visual Question Answering (Video TextVQA), a TEA framework is proposed, based on a T5 generative model, to restore the spatiotemporal relationships between scene text and visual entities in videos, and to aggregate scene text clues related to the question to enhance the quality of answer generation.
TrackGo: A Flexible and Efficient Method for Controllable Video Generation
Haitao Zhou (Beihang University), Changhu Wang (AIsphere Tech)
GenerationData SynthesisDiffusion modelVideo
🎯 What it does: This paper presents TrackGo, a method for precise and controllable video generation using free-form masks and arrows.
Tracking Everything Everywhere across Multiple Cameras
Li-Heng Wang (Academia Sinica), Tyng-Luh Liu (Academia Sinica)
Object TrackingKnowledge DistillationVideo
🎯 What it does: A unified global primitive space is constructed for tracking pixel points in multi-camera videos, enabling incremental learning that maintains pixel correspondence across different times and perspectives.
Trading Off Quality and Uncertainty Through Multi-Objective Optimisation in Batch Bayesian Optimisation
Chao Jiang (University of Birmingham), Miqing Li (University of Birmingham)
Optimization
🎯 What it does: This paper proposes a batch Bayesian optimization method called POEE, which selects batch samples through a dynamically updated quality (posterior mean) and uncertainty (posterior variance) Pareto front, and employs TOPSIS for multi-objective decision-making.
Tradutor: Building a Variety Specific Translation Model
Hugo Sousa (University of Porto), Alipio Jorge (University of Porto)
TransformerSupervised Fine-TuningPrompt EngineeringText
🎯 What it does: A parallel corpus of European Portuguese and English, PTradutor, was created, and an open-source machine translation model was trained based on it.
Traffic Scenario Logic: A Spatial-Temporal Logic for Modeling and Reasoning of Urban Traffic Scenarios
Ruolin Wang (University of Science and Technology of China), Jianmin Ji (University of Science and Technology of China)
Autonomous DrivingGraph
🎯 What it does: This paper proposes and implements a spatial-temporal logic for pedestrian-free urban road traffic scenarios—Traffic Scenario Logic (TSL), achieving automatic conversion from OpenDRIVE high-precision maps to logical models, and generating complete traffic scenario sequences for testing, decision-making, and control validation using the ASP+ temporal logic solver Telingo.
TRAIL: Trust-Aware Client Scheduling for Semi-Decentralized Federated Learning
Gangqiang Hu (Zhejiang Normal University), Hao Fu (Wuhan University of Science and Technology)
OptimizationFederated LearningImage
🎯 What it does: In the semi-distributed federated learning framework, the TRAIL mechanism is proposed, which combines the Adaptive Hidden Semi-Markov Model (AHSMM) to predict client communication and training status, and optimizes client-server association through a greedy algorithm to minimize global training loss.
Training Consistent Mixture-of-Experts-Based Prompt Generator for Continual Learning
Yue Lu (Northwestern Polytechnical University), Yanning Zhang (Xidian University)
ClassificationRecognitionTransformerPrompt EngineeringMixture of ExpertsImage
🎯 What it does: This paper proposes a consistency mixture of experts (MoE) prompt generator to maintain model stability in a continual learning scenario without sample memory;
Training Deep Neural Networks with Virtual Smoothing Classes
Zhiyang Zhou (Chinese Academy of Sciences), Yan Cai (Chinese Academy of Sciences)
ClassificationKnowledge DistillationAdversarial AttackConvolutional Neural NetworkImage
🎯 What it does: A Virtual Smoothing label is proposed, which transforms the overconfidence problem into a competition between the positive class and a virtual class by adding an additional virtual category to the last layer of the classifier, thereby reducing overconfidence while maintaining confidence levels.
Training Matting Models Without Alpha Labels
Wenze Liu (Chinese University of Hong Kong), Xiangyu Yue (Huazhong University of Science and Technology)
RestorationSegmentationTransformerImage
🎯 What it does: Using a rough trimap as training labels, we propose a Direction Distance Consistency loss (DDC loss) to train a deep image matting model without fine alpha annotations.
Training on the Benchmark Is Not All You Need
Shiwen Ni (Shenzhen Institutes of Advanced Technology), Min Yang (Shenzhen Institutes of Advanced Technology)
Anomaly DetectionLarge Language ModelTextBenchmark
🎯 What it does: A leakage detection method based on multiple-choice question option replacement is proposed, and the identification of benchmark test set leakage in LLM pre-training data is achieved under gray-box conditions.
Training Verification-Friendly Neural Networks via Neuron Behavior Consistency
Zongxin Liu (Chinese Academy of Sciences), Lijun Zhang (Chinese Academy of Sciences)
OptimizationAdversarial AttackConvolutional Neural NetworkImage
🎯 What it does: A neural network training method is proposed and implemented, which incorporates neuron behavior consistency (NBC) regularization during the training process to maintain consistent activation states within local neighborhoods, thereby reducing the number of unstable neurons and enhancing the verifiability of the network.
Training with “Paraphrasing the Original Text” Teaches LLM to Better Retrieve in Long-Context Tasks
Yijiong Yu (Tsinghua University), Zhe Zhou (Tsinghua University)
RetrievalTransformerLarge Language ModelSupervised Fine-TuningPrompt EngineeringTextChain-of-Thought
🎯 What it does: This paper proposes a method to explicitly extract key information in long-context tasks by incorporating a 'text rewriting' step into the answers, thereby enhancing the retrieval and question-answering performance of large language models.
Training-and-Prompt-Free General Painterly Harmonization via Zero-Shot Disentenglement on Style and Content References
Teng-Fang Hsiao (National Yang Ming Chiao Tung University), Hong-Han Shuai (National Yang Ming Chiao Tung University)
Image TranslationImage HarmonizationGenerationDiffusion modelImageBenchmark
🎯 What it does: A general painting style fusion method called TF-GPH is proposed, which can achieve seamless integration of foreground and background in three types of tasks: object insertion, object exchange, and style transfer, without the need for training or prompts.
Training-Free and Hardware-Friendly Acceleration for Diffusion Models via Similarity-based Token Pruning
Evelyn Zhang (Shanghai Jiao Tong University), Linfeng Zhang (Shanghai Jiao Tong University)
GenerationComputational EfficiencyDiffusion modelImage
🎯 What it does: This paper proposes and implements SiTo, a similarity-based token pruning method designed to accelerate the inference of diffusion models without the need for training or labeling.
Training-Free Image Manipulation Localization Using Diffusion Models
Zhenfei Zhang (University at Albany State University of New York), Xin Li (University at Albany State University of New York)
Diffusion modelImage
🎯 What it does: A training-free and condition-free diffusion model framework is proposed for locating image tampering.
Training-free Open-Vocabulary Semantic Segmentation via Diverse Prototype Construction and Sub-region Matching
Xuanpu Zhao (University of Science and Technology of China), Nenghai Yu (University of Science and Technology of China)
SegmentationLarge Language ModelDiffusion modelImage
🎯 What it does: A training-free open vocabulary semantic segmentation framework is proposed, achieving fine segmentation through LLM-guided diverse prototype construction and sub-region matching.
Transfer Learning Meets Functional Linear Regression: No Negative Transfer Under Posterior Drift
Xiaoyu Hu (Xi'an Jiaotong University), Zhenhua Lin (National University of Singapore)
Domain AdaptationRecommendation SystemAnomaly DetectionOptimizationFederated LearningData-Centric LearningTabularTime SeriesFinance Related
🎯 What it does: An algorithm is proposed to improve the accuracy of slope function estimation in functional linear regression under the condition of posterior drift through transfer learning;
Transfer Learning of Real Image Features with Soft Contrastive Loss for Fake Image Detection
Ziyou Liang (Wuhan University), Xinyi Yang (Wuhan University)
ClassificationAnomaly DetectionConvolutional Neural NetworkGenerative Adversarial NetworkContrastive LearningImage
🎯 What it does: This paper proposes a Natural Trace Forensics (NTF) framework that detects forged images synthesized by unknown generative models by learning stable features (natural traces) shared by real images, combined with soft contrastive learning and transfer learning.
Transferable Adversarial Face Attack with Text Controlled Attribute
Wenyun Li (Harbin Institute of Technology), Dongmei Jiang (Pengcheng Laboratory)
GenerationAdversarial AttackMeta LearningVision Language ModelGenerative Adversarial NetworkImage
🎯 What it does: A transferable adversarial face attack method TCA2 based on natural language guidance has been developed, utilizing StyleGAN2 to generate realistic attack samples.
Transformer Layers as Painters
Qi Sun (Sakana AI), Llion Jones (Emergence AI)
TransformerLarge Language ModelText
🎯 What it does: This paper explores the similarity and robustness between layers by conducting various experimental variants such as layer skipping, rearrangement, parallel execution, and cyclic execution on a frozen pre-trained Transformer.
TranSplat: Generalizable 3D Gaussian Splatting from Sparse Multi-View Images with Transformers
Chuanrui Zhang (Tsinghua University), Haoqian Wang (E-surfing Vision Technology Co., Ltd)
GenerationDepth EstimationTransformerGaussian SplattingImage
🎯 What it does: A Transformer-based sparse view 3D Gaussian fitting framework called TranSplat is proposed for general sparse view 3D reconstruction without scene-specific optimization.
Transtreaming: Adaptive Delay-aware Transformer for Real-time Streaming Perception
Xiang Zhang (Southeast University), Weiwei Wu (Southeast University)
Object DetectionAutonomous DrivingTransformerVideo
🎯 What it does: This paper proposes Transtreaming, a real-time streaming perception framework that achieves multi-frame prediction through an adaptive delay-aware Transformer to compensate for computational delays.
Treasures in Discarded Weights for LLM Quantization
Hao Yu (Nanjing University), Jianxin Wu
OptimizationTransformerLarge Language ModelText
🎯 What it does: After low-bit quantization of large language models, compensate for the lost weight information through the search space to improve accuracy;
TreeEval: Benchmark-Free Evaluation of Large Language Models through Tree Planning
Xiang Li (East China Normal University), Chao Yang (Shanghai AI Laboratory)
TransformerLarge Language ModelPrompt EngineeringTextBenchmark
🎯 What it does: Proposes TreeEval, a benchmark-free evaluation method for LLMs based on tree planning.
Tri-Ergon: Fine-Grained Video-to-Audio Generation with Multi-Modal Conditions and LUFS Control
Bingliang Li (vivo Mobile Communication Co., Ltd), Yiran Zhong (OpenNLPLab)
GenerationData SynthesisTransformerDiffusion modelVideoMultimodalityAudio
🎯 What it does: The Tri-Ergon framework is proposed to achieve high-quality, controllable synchronous audio synthesis from video.
Trigger3:Refining Query Correction via Adaptive Model Selector
Kepu Zhang (Renmin University of China), Jun Xu (Renmin University of China)
TransformerLarge Language ModelSupervised Fine-TuningText
🎯 What it does: Proposes the Trigger3 framework, which utilizes a three-stage trigger (error correction trigger, LLM trigger, fallback trigger) to achieve adaptive collaborative error correction between small models and large language models.
TrojanDec: Data-free Detection of Trojan Inputs in Self-supervised Learning
Yupei Liu (Pennsylvania State University), Jinyuan Jia (Pennsylvania State University)
RestorationAnomaly DetectionDiffusion modelContrastive LearningImage
🎯 What it does: This paper proposes TrojanDec, a complete solution for Trojan input detection and recovery using black-box queries of the encoder within a self-supervised learning framework.
Truncated Gaussian Policy for Debiased Continuous Control
Ganghun Lee (Seoul National University), Byoung-Tak Zhang (Seoul National University)
Reinforcement Learning
🎯 What it does: This paper proposes the use of truncated Gaussian policies to address the boundary action bias problem in continuous control tasks.
Trust-GRS: A Trustworthy Training Framework for Graph Neural Network Based Recommender Systems Against Shilling Attacks
Lingyu Mu (Institute of Information Engineering, Chinese Academy of Sciences), Zheng Lin (Institute of Information Engineering, Chinese Academy of Sciences)
Recommendation SystemAnomaly DetectionGraph Neural NetworkContrastive LearningGraph
🎯 What it does: This paper proposes Trust-GRS, a two-stage, prior knowledge-free GNN recommendation system training framework designed to detect and suppress fake users injected by spoof attackers.
Trusted Unified Feature-Neighborhood Dynamics for Multi-View Classification
Haojian Huang (University of Hong Kong), Zhongjiang He (TeleAI)
ClassificationGraph Neural NetworkImage
🎯 What it does: The TUNED model is proposed for multi-view classification, integrating the local feature-neighborhood structure of each view with global consistency, and adaptively fusing evidence through selective Markov random fields.
TrustUQA: A Trustful Framework for Unified Structured Data Question Answering
Wen Zhang (Zhejiang University), Huajun Chen (Ant Group)
TransformerLarge Language ModelPrompt EngineeringGraphTabularTime SeriesRetrieval-Augmented Generation
🎯 What it does: This paper proposes TrustUQA, a trustworthy unified structured data question-answering framework that can simultaneously support natural language questions for tables, knowledge graphs, and temporal knowledge graphs. It uses a Condition Graph (CG) to unify the representation of different types of data and employs a two-layer functional query approach (first generating a simplified query using LLM, then converting it into an executable CG query through rules) to achieve efficient reasoning. Additionally, it introduces dynamic demonstration retrieval to enhance prompt quality and improve model performance.
TSDF-Based Efficient Motion-Compensated Temporal Interpolation for 3D Dynamic Sequences
Soowoong Kim (Electronics and Telecommunications Research Institute), Seungjoon Yang (Ulsan National Institute of Science and Technology)
RestorationData SynthesisComputational EfficiencyRecurrent Neural NetworkOptical FlowPoint CloudMesh
🎯 What it does: Real-time motion compensation time interpolation using TSDF volume to achieve frame rate enhancement of 3D dynamic sequences.
TSVC: Tripartite Learning with Semantic Variation Consistency for Robust Image-Text Retrieval
Shuai Lyu (Beijing University of Posts and Telecommunications), Meina Song (Beijing University of Posts and Telecommunications)
RetrievalImageText
🎯 What it does: The TSVC framework is proposed to address the noise correspondence problem in image-text retrieval, employing tri-party collaborative learning, semantic information variation consistency estimation soft labels, and distribution adaptive soft margin loss.
TTA-FedDG: Leveraging Test-Time Adaptation to Address Federated Domain Generalization
Haoyuan Liang (Sun Yat-sen University), Juepeng Zheng (Sun Yat-sen University)
Domain AdaptationFederated LearningContrastive LearningImage
🎯 What it does: The TTA-FedDG framework and FedSPL method are proposed, which enhance the model's generalization ability to unknown clients in federated domain generalization through test-time adaptation, feature reordering mixing, and strong pseudo-label learning.
TTE: Two Tokens Are Enough to Improve Parameter-Efficient Tuning
Jiacheng Ruan (Shanghai Jiao Tong University), Yuzhuo Fu (Shanghai Jiao Tong University)
ClassificationRecognitionTransformerSupervised Fine-TuningImage
🎯 What it does: This paper studies the overfitting problem in Parameter-Efficient Tuning (PET) and proposes the TTE framework, which effectively alleviates overfitting and enhances PET performance by utilizing globally learnable tokens, instance-specific tokens, and Parameter-Free Cross Attention (PFCA) loss.
Tuning-Free Accountable Intervention for LLM Deployment – a Metacognitive Approach
Zhen Tan (Arizona State University), Huan Liu (University of North Carolina at Chapel Hill)
Explainability and InterpretabilityComputational EfficiencyTransformerLarge Language ModelMixture of ExpertsText
🎯 What it does: This paper proposes the CLEAR framework, which implements a metacognitive intervention method that can automatically identify and correct errors in large language models (LLMs) during the inference phase without the need for fine-tuning.
Two Sides of the Same Coin: Learning the Backdoor to Remove the Backdoor
Qi Zhao (Karlsruhe Institute of Technology), Christian Wressnegger (Karlsruhe Institute of Technology)
ClassificationAnomaly DetectionConvolutional Neural NetworkSupervised Fine-TuningImage
🎯 What it does: A training-time defense method called HARVEY is proposed, which eliminates backdoors through adaptive dataset splitting without the need for clean reference data.
Two-stream Beats One-stream: Asymmetric Siamese Network for Efficient Visual Tracking
Jiawen Zhu (Dalian University of Technology), Huchuan Lu (Dalian University of Technology)
Object TrackingConvolutional Neural NetworkVideo
🎯 What it does: This paper proposes an asymmetric Siamese structure called AsymTrack, which achieves efficient visual tracking by calculating the template only once during the initialization phase and converting it into a modulation signal that is unidirectionally injected into the search branch; at the same time, efficient template modulation (ETM) and object perception enhancement (OPE) modules are designed to further improve accuracy.
Two-Timescale Critic-Actor for Average Reward MDPs with Function Approximation
Prashansa Panda (Indian Institute of Science), Shalabh Bhatnagar (Indian Institute of Science)
OptimizationReinforcement LearningSequential
🎯 What it does: This paper proposes and analyzes the first two-time-scale Critic-Actor algorithm for average reward MDPs using linear function approximation, providing proofs of non-asymptotic and asymptotic convergence, and determining the optimal learning rate and sample complexity.
Twofold Debiasing Enhances Fine-Grained Learning with Coarse Labels
Xin-yang Zhao (Nanjing University of Science and Technology), Yazhou Yao (Nanjing University of Science and Technology)
ClassificationRepresentation LearningContrastive LearningImage
🎯 What it does: For the Coarse-to-Fine Few-Shot task, a Twofold Debiasing (TFB) method is proposed, which enhances fine-grained representation through multi-layer feature fusion reconstruction and intermediate layer feature alignment, and calibrates the fine-grained classifier distribution using features from the coarse label training set.
U-KAN Makes Strong Backbone for Medical Image Segmentation and Generation
Chenxin Li (Chinese University of Hong Kong), Yixuan Yuan (Chinese University of Hong Kong)
SegmentationGenerationConvolutional Neural NetworkDiffusion modelImageBiomedical DataMagnetic Resonance Imaging
🎯 What it does: This paper proposes a new framework called U-KAN, which embeds the Kolmogorov-Arnold Network (KAN) into the U-Net backbone for medical image segmentation and generation.
UAWTrack: Universal 3D Single Object Tracking in Adverse Weather
Yuxiang Yang (Hangzhou Dianzi University), Jing Zhang (Wuhan University)
Object TrackingAutonomous DrivingMixture of ExpertsPoint Cloud
🎯 What it does: This paper proposes UAWTrack, a unified model capable of 3D single-object tracking under various weather conditions.
UCF-Crime-DVS: A Novel Event-Based Dataset for Video Anomaly Detection with Spiking Neural Networks
Yuanbin Qian (Ningbo University), Jiafei Wu (The University of Hong Kong)
Anomaly DetectionSpiking Neural NetworkVideo
🎯 What it does: The first video anomaly detection dataset based on event cameras, UCF-Crime-DVS, has been constructed, and a multi-scale fusion framework (MSF) based on Spiking Neural Networks (SNN) has been proposed for weakly supervised video anomaly detection.
UFO: Enhancing Diffusion-Based Video Generation with a Uniform Frame Organizer
Delong Liu (Beijing University of Posts and Telecommunications), Fei Su (Beijing University of Posts and Telecommunications)
GenerationData SynthesisTransformerDiffusion modelVideo
🎯 What it does: The UFO plugin is proposed, utilizing a lightweight non-intrusive adapter to enhance the consistency and quality of video generation based on diffusion models;
Ultra-High Resolution Segmentation via Boundary-Enhanced Patch-Merging Transformer
Haopeng Sun (Chinese Academy of Sciences), Yiqiang Chen (Chinese Academy of Sciences)
SegmentationTransformerImage
🎯 What it does: This paper proposes the Boundary-Enhanced Patch-Merging Transformer (BPT) for semantic segmentation of high-resolution remote sensing images.
Ultra-High-Definition Dynamic Multi-Exposure Image Fusion via Infinite Pixel Learning
Xingchi Chen (Shenzhen Campus of Sun Yat-sen University), Wenqi Ren (State University of New York at Buffalo)
Image TranslationRestorationCompressionComputational EfficiencyImageVideo
🎯 What it does: Proposes a 4K dynamic multi-exposure image fusion method based on infinite pixel learning.
UN-DETR: Promoting Objectness Learning via Joint Supervision for Unknown Object Detection
HaoMiao Liu, Bo Ma (Beijing Institute of Technology)
Object DetectionTransformerImage
🎯 What it does: This paper proposes an unknown object detection framework called UN-DETR based on Transformer, which achieves end-to-end unknown object detection by combining Instance Presence Score (IPS) predictor, one-to-many assignment, unbiased query selection, IPS-guided post-processing, and unsupervised pre-training.
Unaligned Message-Passing and Contextualized-Pretraining for Robust Geo-Entity Resolution
Yuwen Ji (Beihang University), Yue Zhang (Westlake University)
Graph Neural NetworkTransformerLarge Language ModelText
🎯 What it does: This paper proposes a BERT-based encoder that incorporates Unaligned Message Passing (UMP) and Contextual Pre-training (CP) to enhance the robustness of geographic entity matching through neighborhood geographic context.
Uncertainty-Aware Contrastive Learning with Hard Negative Sampling for Code Search Tasks
Han Liu (Shenzhen University), Qin Zhang (Shenzhen University)
RetrievalRepresentation LearningTransformerContrastive LearningText
🎯 What it does: This paper studies the code retrieval task and proposes an uncertainty-aware contrastive learning and hard negative sampling method to enhance the embedding representations of queries and code.
Uncertainty-Aware Global-View Reconstruction for Multi-View Multi-Label Feature Selection
Pingting Hao (Jilin University), Wanfu Gao (Portland State University)
ClassificationOptimizationGraph Neural NetworkMultimodality
🎯 What it does: A global view reconstruction method for uncertainty perception, UGRFS, is proposed for multi-view multi-label feature selection, integrating sample confidence, view relationships, and label information to generate sparse feature weights.
Uncertainty-Aware Self-Training for CTC-Based Automatic Speech Recognition
Eungbeom Kim (Seoul National University), Kyogu Lee (Seoul National University)
RecognitionTransformerAudio
🎯 What it does: This paper proposes a sequence-level uncertainty estimation based on output probabilities in the CTC model and applies it to self-training to enhance the semi-supervised learning performance of end-to-end ASR.
Uncommon Belief in Rationality
Qi Shi (University of Southampton), Pavel Naumov (University of Southampton)
🎯 What it does: A directed graph-based language (RBR graph) is proposed to characterize the structure of 'non-common belief' among agents, along with the corresponding strategy reasoning process.
Uncovering LLM-Generated Code: A Zero-Shot Synthetic Code Detector via Code Rewriting
Tong Ye (Zhejiang University), Wenhai Wang (Zhejiang University)
AI Code AssistantTransformerLarge Language ModelContrastive LearningTextChain-of-Thought
🎯 What it does: A zero-shot synthetic code detection method is proposed, which first rewrites the code to be detected using a large language model (LLM), and then calculates the similarity between the original code and the rewritten code using a code similarity model. If the similarity is high, it is determined to be code generated by the LLM.
Undermining Mental Proof: How AI Can Make Cooperation Harder by Making Thinking Easier
Zachary Wojtowicz (Harvard University), Simon DeDeo (Carnegie Mellon University)
Review/Survey Paper
🎯 What it does: Proposes the concept of 'psychological proof' and explains its mechanism for promoting cooperation in low-trust environments.
Understanding EFX Allocations: Counting and Variants
Tzeh Yuan Neoh (Institute of High Performance Computing Agency for Science Technology and Research), Nicholas Teh (University of Oxford)
Optimization
🎯 What it does: The study presents the minimum quantities for fair allocations of EFX, WEFX, and EFX+ in the case of a limited number of items relative to the number of participants, proves that the counting problem is #P-complete, and proposes various constructions and algorithms to solve specific cases.
Understanding Emotional Body Expressions via Large Language Models
Haifeng Lu (Lanzhou University), Xiping Hu (Beijing Institute of Technology)
RecognitionGenerationGraph Neural NetworkTransformerLarge Language ModelPrompt EngineeringContrastive LearningTextMultimodality
🎯 What it does: This paper proposes a framework for full-body skeleton emotion recognition and text interpretation based on large language models, called EAI-LLM.
Understanding Individual Agent Importance in Multi-Agent System via Counterfactual Reasoning
Jianming Chen (Institute of Software Chinese Academy of Sciences), Fanjiang Xu (Institute of Software Chinese Academy of Sciences)
Robotic IntelligenceReinforcement LearningAgentic AITabularBenchmark
🎯 What it does: Proposed an EM AI method based on adversarial randomization to evaluate the importance of each agent in a multi-agent system at different time steps.
UniDemoiré: Towards Universal Image Demoiréing with Data Generation and Synthesis
Zemin Yang (ShanghaiTech University), Yuexin Ma (ShanghaiTech University)
RestorationGenerationData SynthesisTransformerDiffusion modelImage
🎯 What it does: This paper proposes a unified image de-moiré framework called UniDemoir, which achieves generalization for various moiré patterns through large-scale data generation and synthesis.
UniDet3D: Multi-dataset Indoor 3D Object Detection
Maksim Kolodiazhnyi (Artificial Intelligence Research Institute), Anton Konushin (Artificial Intelligence Research Institute)
Object DetectionTransformerPoint Cloud
🎯 What it does: UniDet3D trains a unified 3D object detection model using joint supervised training on multiple indoor datasets.
Unified Coding for Both Human Perception and Generalized Machine Analytics with CLIP Supervision
Kangsheng Yin (Shenzhen University), Shiqi Wang (City University of Hong Kong)
Object DetectionSegmentationCompressionTransformerAuto EncoderContrastive LearningImageMultimodality
🎯 What it does: This paper proposes a unified and general image coding framework UG-ICM, which utilizes a single bitstream to meet the needs of human visual perception and unknown machine vision analysis through conditional decoding, and achieves self-supervised training via cross-modal supervision provided by CLIP.
Unified Graph Neural Networks Pre-training for Multi-domain Graphs
Mingkai Lin (Nanjing University), Sanglu Lu (Nanjing University)
Domain AdaptationRepresentation LearningGraph Neural NetworkPrompt EngineeringGraph
🎯 What it does: This paper proposes a multi-domain graph pre-training framework MDP-GNN, aimed at unifying the features and structures of different graph domains by learning potential 'meta-domains' to construct a GNN model that can be generalized across various graph tasks.
Unified Knowledge Maintenance Pruning and Progressive Recovery with Weight Recalling for Large Vision-Language Models
Zimeng Wu (Beihang University), Yunhong Wang (Beihang University)
RetrievalCompressionKnowledge DistillationTransformerLarge Language ModelVision Language ModelImageTextMultimodality
🎯 What it does: This paper proposes a unified structured pruning framework for large-scale vision-language models (LVLM) called UKMP, aimed at significantly reducing model parameters and computational load while maintaining zero-shot performance.
UniFORM: Towards Unified Framework for Anomaly Detection on Graphs
Chuancheng Song (Institute of Information Engineering Chinese Academy of Sciences), Yanan Cao (Institute of Information Engineering Chinese Academy of Sciences)
Anomaly DetectionMeta LearningGraph Neural NetworkContrastive LearningGraph
🎯 What it does: A unified self-supervised graph anomaly detection framework called UniFORM is proposed, capable of detecting anomalies at the node, edge, and graph levels simultaneously.
UniMuMo: Unified Text, Music, and Motion Generation
Han Yang (Chinese University of Hong Kong), Chuang Gan (Cisco Research)
GenerationData SynthesisTransformerTextMultimodalityAudio
🎯 What it does: A unified multimodal generation framework called UniMuMo has been constructed, capable of generating combinations of text, music, and actions.
Union Is Strength! Unite the Power of LLMs and MLLMs for Chart Question Answering
Jiapeng Liu (Institute of Information Engineering, Chinese Academy of Sciences), Can Ma (Institute of Information Engineering, Chinese Academy of Sciences)
TransformerLarge Language ModelPrompt EngineeringMultimodalityTabularChain-of-Thought
🎯 What it does: A framework named SYNERGY is proposed, which integrates large language models (LLM) and multimodal large language models (MLLM) to accomplish chart question answering tasks (CQA), achieving more accurate answer generation through phased processing.
UniPCGC: Towards Practical Point Cloud Geometry Compression via an Efficient Unified Approach
Kangli Wang (Peking University), Wei Gao (Peking University)
CompressionPoint Cloud
🎯 What it does: A unified framework UniPCGC is proposed, achieving lossless, lossy, variable rate, and variable complexity for point cloud geometry compression.
UniTR: A Unified Framework for Joint Representation Learning of Trajectories and Road Networks
Jie Zhao (Chongqing University), Yuxuan Liang (Hong Kong University of Science and Technology)
OptimizationRepresentation LearningGraph Neural NetworkTransformerContrastive LearningMultimodalityGraph
🎯 What it does: The UniTR framework is proposed to achieve joint representation learning of road networks and trajectories, utilizing a hierarchical propagation mechanism and three-layer contrastive learning to simultaneously optimize the representations of both types of data within the same embedding space.
Universal Domain Adaptive Object Detection via Dual Probabilistic Alignment
Yuanfan Zheng (Institute of Automation, Chinese Academy of Sciences), Zhen Chen (Chinese University of Hong Kong)
Object DetectionDomain AdaptationImage
🎯 What it does: Proposes the Dual Probabilistic Alignment (DPA) framework, which targets unified domain adaptive object detection by performing probabilistic alignment for global domain private categories and instance-level domain shared categories, and reduces negative transfer through private category constraints.
Universal Features Guided Zero-Shot Category-Level Object Pose Estimation
Wentian Qu (Institute of Software, Chinese Academy of Sciences), Ping Tan (Hong Kong University of Science and Technology)
Object DetectionPose EstimationGraph Neural NetworkDiffusion modelImagePoint Cloud
🎯 What it does: A zero-shot category-level object 6-DOF pose estimation framework based on 2D/3D universal features is proposed, capable of directly inferring poses on unknown categories.
Universal Post-Processing Networks for Joint Optimization of Modules in Task-Oriented Dialogue Systems
Atsumoto Ohashi (Nagoya University), Ryuichiro Higashinaka (Nagoya University)
OptimizationTransformerLarge Language ModelReinforcement LearningText
🎯 What it does: This paper studies a unified post-processing network (UniPPN) designed to jointly optimize the outputs of all modules in task-oriented dialogue systems, thereby improving the task completion rate of the system.
Universality of Real Minimal Complexity Reservoir
Robert Simon Fong (University of Birmingham), Peter Tino (New Mexico State University)
Time Series
🎯 What it does: It is proven that Simple Cycle Reservoirs (SCR) can approximate any time-invariant dynamic filter with decaying memory over the real number field, filling a gap that was previously only known in the complex number field.
Unlearning Concepts in Diffusion Model via Concept Domain Correction and Concept Preserving Gradient
Yongliang Wu (Southeast University), Xu Yang (Nanyang Technological University)
GenerationData SynthesisDiffusion modelImageText
🎯 What it does: This paper proposes a conceptual forgetting framework called DoCo for text-to-image diffusion models, aimed at thoroughly removing sensitive concepts while maintaining the overall performance of the model.
Unleashing the Potential of Large Language Models as Prompt Optimizers: Analogical Analysis with Gradient-based Model Optimizers
Xinyu Tang (Renmin University of China), Ji-Rong Wen (Renmin University of China)
OptimizationTransformerLarge Language ModelPrompt EngineeringTextRetrieval-Augmented Generation
🎯 What it does: This paper proposes a gradient heuristic-based large language model prompt optimizer (GPO) that can continuously improve task prompts through the retrieval of relevant optimization trajectories and generative fine-tuning in multi-step iterations.
Unleashing the Potential of Model Bias for Generalized Category Discovery
Wenbin An (Xi'an Jiaotong University), Ping Chen (University of Massachusetts Boston)
ClassificationRecognitionTransformerLarge Language ModelText
🎯 What it does: Proposes the Self-Debiasing Calibration (SDC) framework, which utilizes the outputs of a pre-trained bias model to dynamically calibrate logits, generating more accurate pseudo-labels, thereby improving the recognition of new categories in the General Category Discovery (GCD) task.
Unleashing the Power of Visual Foundation Models for Generalizable Semantic Segmentation
PeiYuan Tang (Xi'an Jiaotong University), Zijiang James Yang (Singapore Management University)
SegmentationDomain AdaptationAutonomous DrivingConvolutional Neural NetworkSupervised Fine-TuningImage
🎯 What it does: The study utilizes visual foundation models for domain generalization semantic segmentation.
Unleashing the Temporal-Spatial Reasoning Capacity of GPT for Training-Free Audio and Language Referenced Video Object Segmentation
Shaofei Huang (Hefei University of Technology), Si Liu (Meituan)
Object DetectionSegmentationTransformerLarge Language ModelPrompt EngineeringVideoMultimodalityChain-of-ThoughtAudio
🎯 What it does: A training-free Audio-Language-Referenced SAM 2 (AL-Ref-SAM 2) pipeline is proposed to achieve multi-modal video object segmentation (AVS and RVOS) with audio and language cues;
Unlocking Better Closed-Set Alignment Based on Neural Collapse for Open-Set Recognition
Chaohua Li (Nanjing University of Aeronautics and Astronautics), Songcan Chen (Nanjing University of Aeronautics and Astronautics)
ClassificationRecognitionContrastive LearningImage
🎯 What it does: The paper proposes to enhance open set recognition performance by aligning closed set features with the classifier strictly to an ideal simple isometric compact framework through a fixed ETF template and dual ETF loss.
Unlocking the Game: Estimating Games in Möbius Representation for Explanation and High-Order Interaction Detection
Majid Mohammadi (Vrije Universiteit Amsterdam), Annette Ten Teije (Vrije Universiteit Amsterdam)
Explainability and InterpretabilityComputational EfficiencyTextTabular
🎯 What it does: The GEM-FIX method is proposed, which allows for the precise calculation of Shapley values and efficient identification of significant high-order feature interactions by performing regression on the explanation game using Möbius representation.
Unlocking the Potential of Black-box Pre-trained GNNs for Graph Few-shot Learning
Qiannan Zhang (Cornell University), Xiangliang Zhang (University of Notre Dame)
Meta LearningGraph Neural NetworkGraph
🎯 What it does: This study investigates how to achieve few-shot learning for graph nodes in a black-box pre-trained Graph Neural Network (GNN) environment, proposing a lightweight meta-learner to extract task-related knowledge and utilizing sub-networks for rapid adaptation.
Unlocking the Potential of Reverse Distillation for Anomaly Detection
Xinyue Liu (Beihang University), Shuo Zhang (Beijing Jiaotong University)
Anomaly DetectionKnowledge DistillationConvolutional Neural NetworkImage
🎯 What it does: This paper proposes an unsupervised anomaly detection method that incorporates an expert network and guided information injection within a reverse knowledge distillation framework, which can simultaneously enhance the teacher network's sensitivity to anomalies and the student network's denoising capability.
Unlocking the Power of LSTM for Long Term Time Series Forecasting
Yaxuan Kong (University of Oxford), Qingsong Wen (Squirrel AI)
Recurrent Neural NetworkTime Series
🎯 What it does: A long-term forecasting model for time series based on sLSTM is proposed—P-sLSTM, which addresses the short memory problem of traditional LSTM by utilizing patching and channel independence techniques to enhance memory and generalization capabilities.
Unlocking the Power of Patch: Patch-Based MLP for Long-Term Time Series Forecasting
Peiwang Tang (University of Science and Technology of China), Weitai Zhang (University of Science and Technology of China)
Time Series
🎯 What it does: This paper proposes a Patch-based MLP (PatchMLP) model that utilizes patch embedding and linear layers to predict long-period time series.
Unpaired Multi-Domain Histopathology Virtual Staining Using Dual Path Prompted Inversion
Bing Xiong (Shenzhen Institutes of Advanced Technology), Wenjian Qin (Shenzhen Institutes of Advanced Technology)
Image TranslationData SynthesisPrompt EngineeringDiffusion modelImageBiomedical Data
🎯 What it does: This paper proposes a dual-path prompt inversion method for unpaired multi-domain virtual staining, utilizing deterministic inversion from a pre-trained diffusion model and optimizable visual prompts to maintain structural consistency and achieve style-controllable separation.
Unravelling Causal Genetic Biomarkers of Alzheimer’s Disease via Neuron to Gene-token Backtracking in Neural Architecture: A Groundbreaking Reverse-Gene-Finder Approach
Victor O. K. Li (University of Hong Kong), Jacqueline C. K. Lam (University of Hong Kong)
ClassificationDrug DiscoveryTransformerSupervised Fine-TuningBiomedical DataAlzheimer's Disease
🎯 What it does: Based on the large gene pre-training model Geneformer, combined with single-cell microglial expression data, we propose the Reverse-Gene-Finder method, which first fine-tunes the model for early classification of Alzheimer's disease, then identifies the most causative neurons (MCN) through causal tracing, and finally traces back to the input layer to obtain the most causative gene markers (MCT) and their corresponding genes (MCG), thereby discovering several new candidate genes previously unrecognized as AD-related.
Unsupervised Anomaly Detection for Tabular Data Using Deep Noise Evaluation
Wei Dai (Chinese University of Hong Kong), Jicong Fan (Chinese University of Hong Kong)
Anomaly DetectionTabularFinance Related
🎯 What it does: A noise assessment-based unsupervised anomaly detection method is proposed, which directly trains the model using samples of normal data with added noise to learn the anomaly boundary.
Unsupervised Audio-Visual Segmentation with Modality Alignment
Swapnil Bhosale (University of Surrey), Xiatian Zhu (Imperial College London)
SegmentationContrastive LearningVideoMultimodalityAudio
🎯 What it does: A novel unsupervised audio-video segmentation framework MoCA is proposed, utilizing existing foundational models (DINO, ImageBind, SAM) to achieve pixel-level correspondence between audio and visual data.
Unsupervised Degradation Representation Aware Transform for Real-World Blind Image Super-Resolution
Sen Chen (Xidian University), Yaowei Wang (Tianjin University)
RestorationSuper ResolutionContrastive LearningImage
🎯 What it does: This paper designs an unsupervised degradation representation-aware transformation network (DRAT), which filters image content through a dual encoder to accurately extract degradation representations and applies them to feature transformation and global degradation-aware blocks, achieving superior blind image super-resolution.
Unsupervised Diffusion-Based Degradation Modeling for Real-World Super-Resolution
Yuying Chen (Shenzhen Campus of Sun Yat-sen University), Wenqi Ren (Shenzhen Campus of Sun Yat-sen University)
RestorationSuper ResolutionDiffusion modelImage
🎯 What it does: An unsupervised diffusion degradation modeling framework (UDDM) is proposed, which maps real low-resolution images to extremely low resolution through extreme downsampling, learns the degradation distribution, and generates content-aware low-resolution images using a physical dynamic degradation module (P-DDM), ultimately synthesizing HR-LR training pairs consistent with the real distribution for training SISR models.