Shu-Tao Xia

RelayFormer: A Unified Local-Global Attention Framework for Scalable Image and Video Manipulation Localization

SparseEval: Efficient Evaluation of Large Language Models by Sparse Optimization

PromptHub: Enhancing Multi-Prompt Visual In-Context Learning with Locality-Aware Fusion, Concentration and Alignment

Closing the Safety Gap: Surgical Concept Erasure in Visual Autoregressive Models

Revisiting Backdoor Attacks on LLMs: A Stealthy and Practical Poisoning Framework via Harmless Inputs

SP-VLA: A Joint Model Scheduling and Token Pruning Approach for VLA Model Acceleration

Distribution-informed Online Conformal Prediction

IDFSR: Personalized Face Super-Resolution with Identity Decoupling and Fitting

EG3AD: An Efficient Geometry-Aware Encoding Framework for Reconstruction-Based Multi-Class Point Cloud Anomaly Detection

PIC: Revisiting INR for Image Coding with Fast Encoding and Sub-Millisecond Decoding

NextStep-1: Toward Autoregressive Image Generation with Continuous Tokens at Scale

MambaSIC: Mamba-based Stereo Image Compression with Bi-directional Multi-reference Entropy Model

Why does Weak-OOD Help? A Further Step Towards Understanding Jailbreaking VLMs

Imperceptible Jailbreaking against Large Language Models

Grounding Language with Vision: A Conditional Mutual Information Calibrated Decoding Strategy for Reducing Hallucinations in LVLMs

TimeBridge: Non-Stationarity Matters for Long-term Time Series Forecasting

3D-LMVIC: Learning-based Multi-View Image Compression with 3D Gaussian Geometric Priors

Error-quantified Conformal Inference for Time Series

An Exploration with Entropy Constrained 3D Gaussians for 2D Video Compression

IntLoRA: Integral Low-rank Adaptation of Quantized Diffusion Models

TimeFilter: Patch-Specific Spatial-Temporal Graph Filtration for Time Series Forecasting

Stealthy Shield Defense: A Conditional Mutual Information-Based Approach against Black-Box Model Inversion Attacks

DreamBench++: A Human-Aligned Benchmark for Personalized Image Generation

IntLoRA: Integral Low-rank Adaptation of Quantized Diffusion Models

3D-GP-LMVIC: Learning-based Multi-View Image Compression with 3D Gaussian Geometric Priors

Going Beyond Feature Similarity: Effective Dataset distillation based on Class-aware Conditional Mutual Information

Block-to-Scene Pre-training for Point Cloud Hybrid-Domain Masked Autoencoders

One Perturbation is Enough: On Generating Universal Adversarial Perturbations against Vision-Language Pre-training Models

Towards Scalable Semantic Representation for Recommendation

MambaVC: Exploring Selective State Spaces for Learned Visual Compression

TimeBridge: Non-Stationarity Matters for Long-term Time Series Forecasting

CALoR: Towards Comprehensive Model Inversion Defense

MIBench: A Comprehensive Benchmark for Model Inversion Attack and Defense

WATER-GS: Watermark-embedded 3D Gaussian Splatting via Plug-and-play Decoder

Denial-of-Service Poisoning Attacks against Large Language Models

RouGE: Learning Gated Experts for Segment Anything in the Wild