Plug-and-Play Grounding of Reasoning in Multimodal Large Language Models
This addresses a bottleneck in visual reasoning for MLLM applications, offering an incremental improvement over existing methods.
The paper tackles the problem of Multimodal Large Language Models (MLLMs) struggling to capture fine details in high-resolution images by introducing P2G, a plug-and-play framework that uses expert agents for grounding reasoning, achieving performance comparable to GPT-4V on a new benchmark with a 7B backbone.
The rise of Multimodal Large Language Models (MLLMs), renowned for their advanced instruction-following and reasoning capabilities, has significantly propelled the field of visual reasoning. However, due to limitations in their image tokenization processes, most MLLMs struggle to capture fine details of text and objects in images, especially in high-resolution samples. To overcome this limitation, we introduce P2G, a novel framework for plug-and-play grounding in MLLMs. P2G utilizes the tool-usage potential of MLLMs to employ expert agents for on-the-fly grounding of reasoning into critical visual and textual elements in images, thereby enabling deliberate reasoning through multimodal prompting. Additionally, we develop P2GB, a benchmark designed to evaluate MLLMs' proficiency in understanding inter-object relationships and textual content in challenging high-resolution images. Extensive experiments on visual reasoning tasks demonstrate the superiority of P2G, achieving performance comparable to GPT-4V on P2GB with a 7B backbone. Our work underscores the potential of grounding reasoning with external agents in MLLMs, presenting a promising alternative to mere model scaling.