Enhancing Adversarial Training with Prior Knowledge Distillation for Robust Image Compression
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2024
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866914715756658688 |
|---|---|
| author | Cao, Zhi Bao, Youneng Meng, Fanyang Li, Chao Tan, Wen Wang, Genhong Liang, Yongsheng |
| author_facet | Cao, Zhi Bao, Youneng Meng, Fanyang Li, Chao Tan, Wen Wang, Genhong Liang, Yongsheng |
| contents | Deep neural network-based image compression (NIC) has achieved excellent performance, but NIC method models have been shown to be susceptible to backdoor attacks. Adversarial training has been validated in image compression models as a common method to enhance model robustness. However, the improvement effect of adversarial training on model robustness is limited. In this paper, we propose a prior knowledge-guided adversarial training framework for image compression models. Specifically, first, we propose a gradient regularization constraint for training robust teacher models. Subsequently, we design a knowledge distillation based strategy to generate a priori knowledge from the teacher model to the student model for guiding adversarial training. Experimental results show that our method improves the reconstruction quality by about 9dB when the Kodak dataset is elected as the backdoor attack object for psnr attack. Compared with Ma2023, our method has a 5dB higher PSNR output at high bitrate points. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2403_06700 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | Enhancing Adversarial Training with Prior Knowledge Distillation for Robust Image Compression Cao, Zhi Bao, Youneng Meng, Fanyang Li, Chao Tan, Wen Wang, Genhong Liang, Yongsheng Image and Video Processing Deep neural network-based image compression (NIC) has achieved excellent performance, but NIC method models have been shown to be susceptible to backdoor attacks. Adversarial training has been validated in image compression models as a common method to enhance model robustness. However, the improvement effect of adversarial training on model robustness is limited. In this paper, we propose a prior knowledge-guided adversarial training framework for image compression models. Specifically, first, we propose a gradient regularization constraint for training robust teacher models. Subsequently, we design a knowledge distillation based strategy to generate a priori knowledge from the teacher model to the student model for guiding adversarial training. Experimental results show that our method improves the reconstruction quality by about 9dB when the Kodak dataset is elected as the backdoor attack object for psnr attack. Compared with Ma2023, our method has a 5dB higher PSNR output at high bitrate points. |
| title | Enhancing Adversarial Training with Prior Knowledge Distillation for Robust Image Compression |
| topic | Image and Video Processing |
| url | https://arxiv.org/abs/2403.06700 |