Increasing the Robustness of Image Quality Assessment Models Through Adversarial Training
Open Access
- 5 November 2024
- journal article
- research article
- Published by MDPI AG in Technologies
- Vol. 12 (11) , 220
Abstract
The adversarial robustness of image quality assessment (IQA) models to adversarial attacks is emerging as a critical issue. Adversarial training has been widely used to improve the robustness of neural networks to adversarial attacks, but little in-depth research has examined adversarial training as a way to improve IQA model robustness. This study introduces an enhanced adversarial training approach tailored to IQA models; it adjusts the perceptual quality scores of adversarial images during training to enhance the correlation between an IQA model’s quality and the subjective quality scores. We also propose a new method for comparing IQA model robustness by measuring the Integral Robustness Score; this method evaluates the IQA model resistance to a set of adversarial perturbations with different magnitudes. We used our adversarial training approach to increase the robustness of five IQA models. Additionally, we tested the robustness of adversarially trained IQA models to 16 adversarial attacks and conducted an empirical probabilistic estimation of this feature.Keywords
All Related Versions
Funding Information
- research centers in the field of artificial intelligence (000000D730321P5Q0002, 70-2021-00142)
- Ivannikov Institute for System Programming (000000D730321P5Q0002, 70-2021-00142)
This publication has 37 references indexed in Scilit:
- Self-organized underwater image enhancementISPRS Journal of Photogrammetry and Remote Sensing, 2024
- INSPIRATION: A reinforcement learning-based human visual perception-driven image enhancement paradigm for underwater scenesEngineering Applications of Artificial Intelligence, 2024
- Inpainting Semantic and Depth Features to Improve Visual Place Recognition in the WildIEEE Access, 2024
- Towards adversarial robustness verification of no-reference image- and video-quality metricsComputer Vision and Image Understanding, 2023
- On the generation of adversarial examples for image quality assessmentThe Visual Computer, 2023
- A Reinforcement Learning Paradigm of Configuring Visual Enhancement for Object Detection in Underwater ScenesIEEE Journal of Oceanic Engineering, 2023
- Underwater Image Enhancement Method via Multi-Interval Subhistogram Perspective EqualizationIEEE Journal of Oceanic Engineering, 2023
- Temporally coherent person matting trained on fake-motion datasetDigital Signal Processing, 2022
- Blind Natural Image Quality Prediction Using Convolutional Neural Networks And Weighted Spatial Pooling2020
- KADID-10k: A Large-scale Artificially Distorted IQA Database2019