Patch Knowledge Transfer for Efficient AI-Generated Image Quality Assessment
Quick Answer
This paper shows that The Patch Knowledge Transfer (PKT) framework optimizes AI-generated image quality assessment by reducing computational costs by 67.7% while maintaining performance comparable to a teacher model.
Quick Take
The Patch Knowledge Transfer (PKT) framework optimizes AI-generated image quality assessment by reducing computational costs by 67.7% while maintaining performance comparable to a teacher model. This dual-model architecture employs a teacher-student setup for efficient knowledge distillation, addressing the limitations of current methods in real-time inference and accuracy. Extensive tests on four AIGIQA databases confirm its superior balance between efficiency and assessment accuracy.
Key Points
- PKT framework reduces computational costs by 67.7% while maintaining performance.
- Utilizes a dual-model architecture with a teacher providing high-quality supervision.
- Student model inherits representation capacity through multi-level supervision.
- Extensive experiments conducted on four AIGIQA databases validate the approach.
- Achieves a superior balance between model efficiency and assessment accuracy.
Paper Resources
📖 Reader Mode
~2 min readAbstract:With the rapid advancement of image generation technologies, perceptual quality assessment of AI-generated images has emerged as a crucial research direction in computer vision. The core challenge of this task lies in achieving efficient quality assessment for massive generated images. Current mainstream approaches exhibit two key limitations: 1) Methods employing complex feature extraction strategies, while improving performance, incur prohibitive computational costs that hinder real-time inference; 2) Simple image scaling-based solutions, despite their computational efficiency, demonstrate significantly inferior assessment accuracy. To address this critical issue, we propose Patch Knowledge Transfer (PKT), a knowledge distillation-based optimization framework that achieves synergistic optimization of visual representation capability and inference efficiency through an innovative multi-level knowledge transfer mechanism. Specifically, we design a dual-model architecture: a teacher model with local-global hybrid processing provides high-quality supervision signals, while a student model relying solely on global processing efficiently inherits the teacher's representation capacity through multi-level supervision. Extensive experiments conducted on 4 AIGIQA databases demonstrate that the PKT framework enables the student model to maintain performance comparable to the teacher while reducing computational costs by 67.7\%. Furthermore, compared to existing methods, our approach achieves a superior balance between model efficiency and assessment accuracy.
| Comments: | 13 pages. ICME26 Spotlight |
| Subjects: | Computer Vision and Pattern Recognition (cs.CV) |
| Cite as: | arXiv:2607.05605 [cs.CV] |
| (or arXiv:2607.05605v1 [cs.CV] for this version) | |
| https://doi.org/10.48550/arXiv.2607.05605 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Jiquan Yuan [view email]
[v1]
Mon, 6 Jul 2026 20:02:25 UTC (568 KB)
— Originally published at arxiv.org
Want this in your inbox every morning?
Daily brief at your local 8am — bilingual EN/中文, free.
More from arXiv cs.CV
See more →LLM-Guided ANN Index Optimization for Human-Object Interaction Retrieval
A phase-aware LLM agent optimizes human-object interaction retrieval, outperforming Optuna TPE by 33.3% and VDTuner by 34.2% on the HICO-DET benchmark. This method enhances throughput by 15.3x over UniIR and demonstrates strong transferability across vector database management systems.