notesum.ai
Published at November 26Grounding-IQA: Multimodal Language Grounding Model for Image Quality Assessment
cs.CV
Released Date: November 26, 2024
Authors: Zheng Chen1, Xun Zhang1, Wenbo Li2, Renjing Pei2, Fenglong Song2, Xiongkuo Min1, Xiaohong Liu1, Xin Yuan3, Yong Guo4, Yulun Zhang1
Aff.: 1Shanghai Jiao Tong University; 2Huawei Noah's Ark Lab; 3Westlake University; 4Max Planck Institute for Informatics

| Method | mIoU | Tag-Recall | BLEU@4 | LLM-Score |
| Baseline | N/A | N/A | 3.62 | 48.25 |
| Raw-Box | 0.5624 | 0.5045 | 20.97 | 61.00 |
| Ref-Box | 0.5851 | 0.5497 | 23.67 | 61.75 |