ChatSpot: Bootstrapping Multimodal LLMs via Precise Referring Instruction Tuning

TL;DR

ChatSpot introduces precise referring instructions and a unified multimodal LLM that supports diverse interactive inputs, enhancing the accuracy and flexibility of human-AI multimodal interactions.

Contribution

The paper presents a novel multimodal LLM, ChatSpot, capable of handling various interactive inputs with precise region referencing, advancing multimodal human-AI interaction capabilities.

Findings

ChatSpot achieves promising performance in region recognition.

The model supports diverse interactive forms like clicks, drag-and-drop, and drawing.

A new multi-grained vision-language dataset was constructed for training and evaluation.

Abstract

Human-AI interactivity is a critical aspect that reflects the usability of multimodal large language models (MLLMs). However, existing end-to-end MLLMs only allow users to interact with them through language instructions, leading to the limitation of the interactive accuracy and efficiency. In this study, we present precise referring instructions that utilize diverse reference representations such as points and boxes as referring prompts to refer to the special region. This enables MLLMs to focus on the region of interest and achieve finer-grained interaction. Based on precise referring instruction, we propose ChatSpot, a unified end-to-end multimodal large language model that supports diverse forms of interactivity including mouse clicks, drag-and-drop, and drawing boxes, which provides a more flexible and seamless interactive experience. We also construct a multi-grained vision-language instruction-following dataset based on existing datasets and GPT-4 generating. Furthermore, we design a series of evaluation tasks to assess the effectiveness of region recognition and interaction. Experimental results showcase ChatSpot's promising performance.