Research map / Vision-language and multimodal / CLIP and zero-shot
Few-Shot Image Quality Assessment via Adaptation of Vision-Language Models
Xudong Li, Zihao Huang, Yan Zhang, Yunhang Shen, Ke Li, Xiawu Zheng, Liujuan Cao, Rongrong Ji
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add meta-learning to prompt pre-training (ablation) | GRMP-IQA · LIVEC (50 labels) | SRCC: 0.788 -> 0.828 | Improved |
| add meta-learning to prompt pre-training (ablation) | GRMP-IQA · KonIQ (50 labels) | SRCC: 0.761 -> 0.811 | Improved |
| pre-train with ERM instead of no pre-training (ablation) | CLIP · LIVEC (50 labels) | SRCC: 0.796 -> 0.788 | Hurt |
| add QGR gradient regularization (ablation) | GRMP-IQA · LIVEC (50 labels) | SRCC: 0.828 -> 0.836 | Improved |
| add QGR gradient regularization (ablation) | GRMP-IQA · KonIQ (50 labels) | SRCC: 0.811 -> 0.853 | Improved |
| add QGR gradient regularization (ablation) | GRMP-IQA · LIVEC (100 labels) | SRCC: 0.848 -> 0.857 | Improved |
| add QGR gradient regularization (ablation) | GRMP-IQA · KonIQ (100 labels) | SRCC: 0.840 -> 0.872 | Improved |
| raise QGR weight lambda from 1 to 5 (ablation) | GRMP-IQA · KonIQ (50 labels) | SRCC: 0.812 -> 0.853 | Improved |
| raise QGR weight lambda from 5 to 7 (ablation) | GRMP-IQA · KonIQ (50 labels) | SRCC: 0.853 -> 0.851 | No clear effect |
| add visual meta-prompt to meta text prompt (ablation) | GRMP-IQA · LIVEC (zero-shot) | SRCC: 0.689 -> 0.770 | Improved |
| add text meta-prompt to meta visual prompt (ablation) | GRMP-IQA · LIVEC (zero-shot) | SRCC: 0.742 -> 0.770 | Improved |
| add meta-learning to text and visual prompts (ablation) | GRMP-IQA · LIVEC (zero-shot) | SRCC: 0.709 -> 0.770 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- Q-Align — Q-Align: Teaching LMMs for Visual Scoring via Discrete Text-Defined Levels (ICML 2024) · better on 1 of 1 shared result
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Boosting Image Quality Assessment through Efficient Transformer Adaptation with Local Feature Enhancement (CVPR 2024 · similarity 0.89)
- Towards Syn-to-Real IQA: A Novel Perspective on Reshaping Synthetic Data Distributions (NeurIPS 2025 · similarity 0.88)
- Toward Generalized Image Quality Assessment: Relaxing the Perfect Reference Quality Assumption (CVPR 2025 · similarity 0.88)
- Adaptive Image Quality Assessment via Teaching Large Multimodal Model to Compare (NeurIPS 2024 · similarity 0.88)
- Grounding-IQA: Grounding Multimodal Language Model for Image Quality Assessment (ICLR 2026 · similarity 0.88)
- Revisiting MLLM Based Image Quality Assessment: Errors and Remedy (AAAI 2026 · similarity 0.87)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
