Hui Wang
6 indexed papers
Publications per year
Top categories
Frequent co-authors
Research Timeline
This paper introduces TC-UMIA, a novel tri-class membership inference attack, demonstrating that machine unlearning can leak privacy risks to the retained data set, and evaluates defense mechanisms to mitigate this risk.
The paper introduces CardioLens, a rigorous evaluation testbed for multi-sequence Cardiac MRI, which reveals that current Multimodal Large Language Models (MLLMs) exhibit a significant 'clinical reality gap' and perform poorly when simulating real-world cardiac interpretation workflows.
The paper introduces CardioLens, a rigorous evaluation testbed for multi-sequence Cardiac MRI, which reveals that current Multimodal Large Language Models (MLLMs) exhibit a significant 'clinical reality gap' and perform poorly when simulating real-world cardiac interpretation workflows.
The paper proposes a framework for multimodal learning under missing modalities, called Latent World Recovery (LWR), which aligns modality-specific embeddings in a shared latent space and fuses available modalities for robust prediction.
The paper proposes a method called self-guidance to improve the reconstruction fidelity of neural speech codecs based on Vector-Quantized VAEs.
This paper introduces Omni Demand Understanding (ODU), a benchmark for evaluating AI assistants' ability to infer user demands from multimodal interaction.
Papers
Omni Demand Understanding: A Benchmark for Contextual User-Intent Inference in Multimodal Interaction
Qi Chen, Yunfei Chu, Haolin He, Yifan Yang +14 more
This paper introduces Omni Demand Understanding (ODU), a benchmark for evaluating AI assistants' ability to infer user demands from multimodal interaction.