Yiming Li

16

Papers

78

Total Citations

Papers (16)

Towards Faithful XAI Evaluation via Generalization-Limited Backdoor Watermark

Multiagent Multitraversal Multimodal Self-Driving: Open MARS Dataset

KVTuner: Sensitivity-Aware Layer-Wise Mixed-Precision KV Cache Quantization for Efficient and Nearly Lossless LLM Inference

Adversarial Attacks against Closed-Source MLLMs via Feature Optimal Alignment

Which Model Generated This Image? A Model-Agnostic Approach for Origin Attribution

Towards Resilient Safety-driven Unlearning for Diffusion Models against Downstream Fine-tuning

Adversarial Exploitation of Data Diversity Improves Visual Localization

IBD-PSC: Input-level Backdoor Detection via Parameter-oriented Scaling Consistency

SleeperMark: Towards Robust Watermark against Fine-Tuning Text-to-image Diffusion Models

Purifying Quantization-conditioned Backdoors via Layer-wise Activation Correction with Distribution Approximation

Extrapolated Urban View Synthesis Benchmark

Unraveling the Effects of Synthetic Data on End-to-End Autonomous Driving

Recovering Parametric Scenes from Very Few Time-of-Flight Pixels

Multi-Region Text-Driven Manipulation of Diffusion Imagery

Not All Prompts Are Secure: A Switchable Backdoor Attack Against Pre-trained Vision Transfomers

Nearest is Not Dearest: Towards Practical Defense against Quantization-conditioned Backdoor Attacks