Reproduce “A Comprehensive Study of Multimodal Large Language Models for Image Quality Assessment” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Tianhe Wu, Kede Ma, Jie Liang, Yujiu Yang, Lei Zhang
This paper evaluates multimodal large language models for image quality assessment by combining three image-presentation formats with three prompting strategies, then selecting challenging human-rated samples using expert-model proxies. GPT-4V reaches an SRCC of 0.881 on SPAQ with double-stimulus chain-of-thought prompting, versus 0.834 with standard prompting. However, color-difference judgments remain weak, limiting generality.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.