VILA-U: a Unified Foundation Model Integrating Visual Understanding and Generation

Yecheng Wu, Zhuoyang Zhang, Junyu Chen +9

VILA-U is a Unified foundation model that integrates Video, Image, Language understanding and generation. Traditional visual language models (VLMs) use separate modules for underst…