Transferring From Distortion to Perception-Oriented Optimization: Just-Noticeable-Distortion-Based Domain Adaptation
- 2 September 2025
- journal article
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Multimedia
- Vol. 27 (15209210) , 8199-8211
Abstract
The perception-distortion- tradeoff reveals the limitation of current low-level deep learning paradigms, i.e., minimizing reconstruction distortion does not guarantee improved perceptual quality. Acknowledging the lack of a reliable perception-oriented optimization function, we are motivated to explore a flexible approach for enhancing perceptual quality by steering the tradeoff to prioritize perception. To this end, we reconsider the perception-distortion function by incorporating the Just-Noticeable-Distortion (JND) mechanism. We mathematically demonstrate that in the common image restoration process, altering the optimization target from natural images to distorted images—where the distortion intensity is constrained by the JND threshold and the distortion type aligns with that arising from the restorer itself—effectively obtained improved perception indices without any changes to the restorer or optimization function. Accordingly, to facilitate various low-level learning models, we are motivated to construct the first large-scale CNN-oriented JND image dataset. Our dataset comprises 500 natural images and 4,500 degraded versions generated by a series of autoencoders, as well as the actual JND judgment results collected through rigorous subjective testing from twenty volunteers. Finally, a learning-based JND inference model is established on the proposed dataset and employed in the proposed JND-based adaptation scheme, where the inferred JND images serve as pseudo-ground truth for the training or fine-tuning processes of low-level vision models. Extensive experiments on image super-resolution and end-to-end image compression across multiple models have shown encouraging improvements in perceptual quality, demonstrating the effectiveness of the proposed scheme. Our dataset is available at: https://github.com/ohq17/CNN-Oriented-JND-Dataset.Keywords
Funding Information
- Basic and Frontier Research Project of PCL
- Major Key Project of PCL
- Basic and Applied Basic Research Foundation of Guangdong Province (2024A1515010454)
- Open Research Fund from Guangdong Laboratory of Artificial Intelligence and Digital Economy (GML-KF-24-27)
This publication has 54 references indexed in Scilit:
- Analyzing Perception-Distortion Tradeoff Using Enhanced Perceptual Super-Resolution Network2019
- Image Super-Resolution Using Very Deep Residual Channel Attention Networks2018
- Survey of visual just noticeable difference estimationFrontiers of Computer Science, 2018
- VideoSet: A large-scale compressed video quality dataset based on JND measurementJournal of Visual Communication and Image Representation, 2017
- Learning a no-reference quality metric for single-image super-resolutionComputer Vision and Image Understanding, 2017
- Perceptually-lossless image coding based on foveated-JND and H.265/HEVCJournal of Visual Communication and Image Representation, 2016
- Ultra-Resolving Face Images by Discriminative Generative Networks2016
- U-Net: Convolutional Networks for Biomedical Image Segmentation2015
- ImageNet Large Scale Visual Recognition ChallengeInternational Journal of Computer Vision, 2015
- Just noticeable distortion model and its applications in video codingSignal Processing: Image Communication, 2005