用CLIP指导小模型做图像质量评估，更准且更轻量。
Visual-Language Model Knowledge Distillation Method for Image Quality Assessment
- 用分级提示词让CLIP输出评分，提升其质量评估能力。
- 小模型通过知识蒸馏获得大模型能力，参数减少40%以上。
- 适合资源有限但需高精度图像质量评估的场景。
图像质量评估（IQA）是计算机视觉的核心任务。基于视觉语言模型（如CLIP）的多模态方法在IQA任务中表现出卓越的泛化能力。针对CLIP在IQA中参数冗余、难以捕捉局部失真特征的问题，本文提出一种视觉语言模型知识蒸馏方法，利用CLIP的IQA知识指导架构更优的学生模型训练。首先设计质量分级提示模板，引导CLIP输出质量分数；随后对CLIP进行微调以增强其IQA能力；最后提出一种模态自适应知识蒸馏策略，实现从CLIP教师模型到学生模型的有效指导。在多个IQA数据集上的实验表明，该方法显著降低模型复杂度，同时优于现有IQA方法，在实际部署中展现出强潜力。
原文摘要 · Abstract (English)
Image Quality Assessment (IQA) is a core task in computer vision. Multimodal methods based on vision-language models, such as CLIP, have demonstrated exceptional generalization capabilities in IQA tasks. To address the issues of excessive parameter burden and insufficient ability to identify local distorted features in CLIP for IQA, this study proposes a visual-language model knowledge distillation method aimed at guiding the training of models with architectural advantages using CLIP's IQA knowledge. First, quality-graded prompt templates were designed to guide CLIP to output quality scores. Then, CLIP is fine-tuned to enhance its capabilities in IQA tasks. Finally, a modality-adaptive knowledge distillation strategy is proposed to achieve guidance from the CLIP teacher model to the student model. Our experiments were conducted on multiple IQA datasets, and the results show that the proposed method significantly reduces model complexity while outperforming existing IQA methods, demonstrating strong potential for practical deployment.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。