This repository contains the model of the paper ViSpeak: Visual Instruction Feedback in Streaming Videos. Code: https://github.com/HumanMLLM/ViSpeak