SAMA: Towards Multi-Turn Referential Grounded Video Chat with Large Language Models

Open in new window