Xiangyu Zhang

23

Papers

384

Total Citations

Papers (23)

DreamLLM: Synergistic Multimodal Comprehension and Creation

Ross3D: Reconstructive Visual Instruction Tuning with 3D-Awareness

Taming Teacher Forcing for Masked Autoregressive Video Generation

CO-SPY: Combining Semantic and Pixel Features to Detect Synthetic Images by AI

Compound Text-Guided Prompt Tuning via Image-Adaptive Cues

Glad: A Streaming Scene Generator for Autonomous Driving

ProSec: Fortifying Code LLMs with Proactive Security Alignment

Perception in Reflection

Temporal Logic-Based Multi-Vehicle Backdoor Attacks against Offline RL Agents in End-to-end Autonomous Driving

TAI3: Testing Agent Integrity in Interpreting User Intent

NeurIPS 2025arXiv

DDAE: Towards Deep Dynamic Vision BERT Pretraining

SAMR: A Spatial-Augmented Mixed Reality Method for Enhancing Vision-Language Models in 3D Scene Understanding

Reflected Flow Matching

Holistic Tokenizer for Autoregressive Image Generation

JailbreakDiffBench: A Comprehensive Benchmark for Jailbreaking Diffusion Models

UNIT: Backdoor Mitigation via Automated Neural Distribution Tightening

SubjectDrive: Scaling Generative Data in Autonomous Driving via Subject Control

Language Prompt for Autonomous Driving

Elijah: Eliminating Backdoors Injected in Diffusion Models via Distribution Shift

Panacea: Panoramic and Controllable Video Generation for Autonomous Driving

LAMP: Learn A Motion Pattern for Few-Shot Video Generation

LOTUS: Evasive and Resilient Backdoor Attacks through Sub-Partitioning

BadPart: Unified Black-box Adversarial Patch Attacks against Pixel-wise Regression Tasks