LMEye: An Interactive Perception Network for Large Language Models

yunxin li, Baotian Hu, Xinyu Chen, Lin Ma, Yong Xu, Min zhang

Primary Area: representation learning for computer vision, audio, language, and other modalities

Code Of Ethics: I acknowledge that I and all co-authors of this work have read and commit to adhering to the ICLR Code of Ethics.

Keywords: Multimodal Large Language Model, Interactive Perception Network, LLMs-based Multimodal Agent

Submission Guidelines: I certify that this submission complies with the submission instructions as described on https://iclr.cc/Conferences/2024/AuthorGuide.

Abstract: Training a Multimodal Large Language Model (MLLM) from scratch, like GPT-4, is resource-intensive. Regarding Large Language Models (LLMs) as the core processor for multimodal information, our paper introduces LMEye, a human-like eye with a play-and-plug interactive perception network, designed to enable dynamic interaction between LLMs and external vision information. Previous methods incorporate visual information into LLMs with a simple visual mapping network or Q-former from BLIP-2. Such networks project the image feature once yet do not consider the interaction between the image and the human input query. Hence, the obtained visual information without being connected to human intention may be inadequate for LLMs to generate intention-following responses, which we refer to as static visual information. LMEye addresses this issue by allowing the LLM to request the desired visual information aligned with various human instructions, which we term as the dynamic visual information interaction. Specifically, LMEye consists of a simple visual mapping network to provide the basic perception of an image for LLMs. It also contains additional modules responsible for acquiring requests from LLMs, performing request-based visual information interaction, and transmitting the resulting interacted visual information to LLMs, respectively. In this way, LLMs act to understand the human query, deliver the corresponding request to the request-based visual information interaction module, and generate the response based on the interleaved multimodal information. We evaluate LMEye through extensive experiments on some multimodal benchmarks, demonstrating that it significantly improves the zero-shot performance on various multimodal tasks compared to previous methods, with less parameters.

Anonymous Url: I certify that there is no URL (e.g., github page) that could be used to find authors' identity.

No Acknowledgement Section: I certify that there is no acknowledgement section in this submission for double blind review.

Submission Number: 5180