state-of-the-art methods
The most advanced and effective techniques or approaches currently available in a given field of AI, often representing the peak of research and development.
- A Driving-Style-Adaptive Framework for Vehicle Trajectory Prediction
- A Dynamic Learning Strategy for Dempster-Shafer Theory with Applications in Classification and Enhancement
- Accurate KV Cache Eviction via Anchor Direction Projection for Efficient LLM Inference
- Adaptable Safe Policy Learning from Multi-task Data with Constraint Prioritized Decision Transformer
- Alleviating Hallucinations in Large Language Models through Multi-Model Contrastive Decoding and Dynamic Hallucination Detection
- Angular Constraint Embedding via SpherePair Loss for Constrained Clustering
- AuroRA: Breaking Low-Rank Bottleneck of LoRA with Nonlinear Mapping
- Automatic Synthetic Data and Fine-grained Adaptive Feature Alignment for Composed Person Retrieval
- BackdoorDM: A Comprehensive Benchmark for Backdoor Learning on Diffusion Model
- BecomingLit: Relightable Gaussian Avatars with Hybrid Neural Shading
- Beyond Value Functions: Single-Loop Bilevel Optimization under Flatness Conditions
- Breaking the Discretization Barrier of Continuous Physics Simulation Learning
- COME: Adding Scene-Centric Forecasting Control to Occupancy World Model
- Cancer Survival Analysis via Zero-shot Tumor Microenvironment Segmentation on Low-resolution Whole Slide Pathology Images
- Complete Structure Guided Point Cloud Completion via Cluster- and Instance-Level Contrastive Learning
- Continual Optimization with Symmetry Teleportation for Multi-Task Learning
- Cooperative Retrieval-Augmented Generation for Question Answering: Mutual Information Exchange and Ranking by Contrasting Layers
- DISCOVER: Automated Curricula for Sparse-Reward Reinforcement Learning
- DualFocus: Depth from Focus with Spatio-Focal Dual Variational Constraints
- Dynamic Gaussian Splatting from Defocused and Motion-blurred Monocular Videos
- Efficient Knowledge Transfer in Federated Recommendation for Joint Venture Ecosystem
- Efficient RAW Image Deblurring with Adaptive Frequency Modulation
- Efficient Rectified Flow for Image Fusion
- Enhancing Consistency of Flow-Based Image Editing through Kalman Control
- Enhancing Contrastive Learning with Variable Similarity
- FRN: Fractal-Based Recursive Spectral Reconstruction Network
- Fair Representation Learning with Controllable High Confidence Guarantees via Adversarial Inference
- Faithful Group Shapley Value
- Fast Computation and Optimization for Opinion-Based Quantities of Friedkin-Johnsen Model
- Fast Projection-Free Approach (without Optimization Oracle) for Optimization over Compact Convex Set
- FedGPS: Statistical Rectification Against Data Heterogeneity in Federated Learning
- FlexEvent: Towards Flexible Event-Frame Object Detection at Varying Operational Frequencies
- FlexWorld: Progressively Expanding 3D Scenes for Flexible-View Exploration
- FrameShield: Adversarially Robust Video Anomaly Detection
- From Dormant to Deleted: Tamper-Resistant Unlearning Through Weight-Space Regularization
- GRIFFIN: Effective Token Alignment for Faster Speculative Decoding
- GUI-Actor: Coordinate-Free Visual Grounding for GUI Agents
- Gated Integration of Low-Rank Adaptation for Continual Learning of Large Language Models
- Generalizable Reasoning through Compositional Energy Minimization
- HAODiff: Human-Aware One-Step Diffusion via Dual-Prompt Guidance
- HouseLayout3D: A Benchmark and Training-free Baseline for 3D Layout Estimation in the Wild
- Hybrid Re-matching for Continual Learning with Parameter-Efficient Tuning
- Hyper-Modality Enhancement for Multimodal Sentiment Analysis with Missing Modalities
- IA-GGAD: Zero-shot Generalist Graph Anomaly Detection via Invariant and Affinity Learning
- IOSTOM: Offline Imitation Learning from Observations via State Transition Occupancy Matching
- Image Editing As Programs with Diffusion Models
- Inverse Methods for Missing Data Imputation
- Iterative Foundation Model Fine-Tuning on Multiple Rewards
- Joint Modeling of fMRI and EEG Imaging Using Ordinary Differential Equation-Based Hypergraph Neural Networks
- Learning Pattern-Specific Experts for Time Series Forecasting Under Patch-level Distribution Shift
- Learning Provably Improves the Convergence of Gradient Descent
- Learning from Videos for 3D World: Enhancing MLLMs with 3D Vision Geometry Priors
- MUVR: A Multi-Modal Untrimmed Video Retrieval Benchmark with Multi-Level Visual Correspondence
- MiCADangelo: Fine-Grained Reconstruction of Constrained CAD Models from 3D Scans
- Mitra: Mixed Synthetic Priors for Enhancing Tabular Foundation Models
- Multimodal 3D Genome Pre-training
- NeuroH-TGL: Neuro-Heterogeneity Guided Temporal Graph Learning Strategy for Brain Disease Diagnosis
- OCN: Effectively Utilizing Higher-Order Common Neighbors for Better Link Prediction
- Object-Centric Representation Learning for Enhanced 3D Semantic Scene Graph Prediction
- Offline RL by Reward-Weighted Fine-Tuning for Conversation Optimization
- Omnidirectional 3D Scene Reconstruction from Single Image
- One Sample is Enough to Make Conformal Prediction Robust
- Open-Vocabulary Part Segmentation via Progressive and Boundary-Aware Strategy
- PlanarGS: High-Fidelity Indoor 3D Gaussian Splatting Guided by Vision-Language Planar Priors
- Pragmatic Heterogeneous Collaborative Perception via Generative Communication Mechanism
- Problem-Parameter-Free Decentralized Bilevel Optimization
- Prompt-guided Disentangled Representation for Action Recognition
- Pruning Spurious Subgraphs for Graph Out-of-Distribution Generalization
- PyraMotion: Attentional Pyramid-Structured Motion Integration for Co-Speech 3D Gesture Synthesis
- Q-Insight: Understanding Image Quality via Visual Reinforcement Learning
- Quantifying Distributional Invariance in Causal Subgraph for IRM-Free Graph Generalization
- Reinforced Active Learning for Large-Scale Virtual Screening with Learnable Policy Model
- RepLDM: Reprogramming Pretrained Latent Diffusion Models for High-Quality, High-Efficiency, High-Resolution Image Generation
- Repurposing AlphaFold3-like Protein Folding Models for Antibody Sequence and Structure Co-design
- Repurposing Marigold for Zero-Shot Metric Depth Estimation via Defocus Blur Cues
- STAR: A Benchmark for Astronomical Star Fields Super-Resolution
- Salient Concept-Aware Generative Data Augmentation
- Sample-Efficient Multi-Round Generative Data Augmentation for Long-Tail Instance Segmentation
- SegMASt3R: Geometry Grounded Segment Matching
- Semi-supervised Graph Anomaly Detection via Robust Homophily Learning
- Sequential Multi-Agent Dynamic Algorithm Configuration
- SingRef6D: Monocular Novel Object Pose Estimation with a Single RGB Reference
- Spectral Graph Coarsening Using Inner Product Preservation and the Grassmann Manifold
- Struct-Bench: A Benchmark for Differentially Private Structured Text Generation
- SurfelSplat: Learning Efficient and Generalizable Gaussian Surfel Representations for Sparse-View Surface Reconstruction
- Test-Time Adaptation by Causal Trimming
- Test-Time Spectrum-Aware Latent Steering for Zero-Shot Generalization in Vision-Language Models
- Test3R: Learning to Reconstruct 3D at Test Time
- Towards Generalizable Multi-Policy Optimization with Self-Evolution for Job Scheduling
- Towards Implicit Aggregation: Robust Image Representation for Place Recognition in the Transformer Era
- Track, Inpaint, Resplat: Subject-driven 3D and 4D Generation with Progressive Texture Infilling
- Uncertainty Quantification for Deep Regression using Contextualised Normalizing Flows
- UniMRSeg: Unified Modality-Relax Segmentation via Hierarchical Self-Supervised Compensation
- UniSite: The First Cross-Structure Dataset and Learning Framework for End-to-End Ligand Binding Site Detection
- UrbanIng-V2X: A Large-Scale Multi-Vehicle, Multi-Infrastructure Dataset Across Multiple Intersections for Cooperative Perception
- Vulnerable Data-Aware Adversarial Training
- WarpGAN: Warping-Guided 3D GAN Inversion with Style-Based Novel View Inpainting
- WeatherPrompt: Multi-modality Representation Learning for All-Weather Drone Visual Geo-Localization
- \(\varepsilon\)-Optimally Solving Two-Player Zero-Sum POSGs
- xLSTM-Mixer: Multivariate Time Series Forecasting by Mixing via Scalar Memories