Towards Unified Vision-Language Models with Incomplete Multi-Modal Inputs

Xiang Fang, Wanlong Fang, Changshuo Wang +4

Video-Language Models (VLMs) have demonstrated impressive multi-modal reasoning capabilities across diverse computer vision applications. However, these VLMs are task-specific and…