Xianghu Yue

CoAVT integrates audio, visual, and text data for enhanced understanding.

A new method improves sound localization accuracy while ensuring data privacy.

Introducing a new model and benchmark for evaluating multi-audio tasks.