Efficient Multimodal Large Language Models: A Survey
It tackles the efficiency bottleneck for researchers and practitioners in academia and industry, especially in edge computing, but is incremental as it surveys existing work.
This survey addresses the problem of high computational costs hindering the widespread use of Multimodal Large Language Models (MLLMs) by reviewing efficient and lightweight approaches, summarizing timelines, structures, strategies, and applications.
In the past year, Multimodal Large Language Models (MLLMs) have demonstrated remarkable performance in tasks such as visual question answering, visual understanding and reasoning. However, the extensive model size and high training and inference costs have hindered the widespread application of MLLMs in academia and industry. Thus, studying efficient and lightweight MLLMs has enormous potential, especially in edge computing scenarios. In this survey, we provide a comprehensive and systematic review of the current state of efficient MLLMs. Specifically, we summarize the timeline of representative efficient MLLMs, research state of efficient structures and strategies, and the applications. Finally, we discuss the limitations of current efficient MLLM research and promising future directions. Please refer to our GitHub repository for more details: https://github.com/lijiannuist/Efficient-Multimodal-LLMs-Survey.