Enregistré dans:
| Auteurs principaux: | , , , , , , , , , |
|---|---|
| Format: | Preprint |
| Publié: |
2023
|
| Sujets: | |
| Accès en ligne: | https://arxiv.org/abs/2310.03661 |
| Tags: |
Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
|
| _version_ | 1866909114187120640 |
|---|---|
| author | Bai, Jianhong Yang, Yuchen Chu, Huanpeng Wang, Hualiang Liu, Zuozhu Chen, Ruizhe He, Xiaoxuan Mu, Lianrui Cai, Chengfei Hu, Haoji |
| author_facet | Bai, Jianhong Yang, Yuchen Chu, Huanpeng Wang, Hualiang Liu, Zuozhu Chen, Ruizhe He, Xiaoxuan Mu, Lianrui Cai, Chengfei Hu, Haoji |
| contents | Quantization has emerged as a promising direction for model compression. Recently, data-free quantization has been widely studied as a promising method to avoid privacy concerns, which synthesizes images as an alternative to real training data. Existing methods use classification loss to ensure the reliability of the synthesized images. Unfortunately, even if these images are well-classified by the pre-trained model, they still suffer from low semantics and homogenization issues. Intuitively, these low-semantic images are sensitive to perturbations, and the pre-trained model tends to have inconsistent output when the generator synthesizes an image with poor semantics. To this end, we propose Robustness-Guided Image Synthesis (RIS), a simple but effective method to enrich the semantics of synthetic images and improve image diversity, further boosting the performance of downstream data-free compression tasks. Concretely, we first introduce perturbations on input and model weight, then define the inconsistency metrics at feature and prediction levels before and after perturbations. On the basis of inconsistency on two levels, we design a robustness optimization objective to enhance the semantics of synthetic images. Moreover, we also make our approach diversity-aware by forcing the generator to synthesize images with small correlations in the label space. With RIS, we achieve state-of-the-art performance for various settings on data-free quantization and can be extended to other data-free compression tasks. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2310_03661 |
| institution | arXiv |
| publishDate | 2023 |
| record_format | arxiv |
| spellingShingle | Robustness-Guided Image Synthesis for Data-Free Quantization Bai, Jianhong Yang, Yuchen Chu, Huanpeng Wang, Hualiang Liu, Zuozhu Chen, Ruizhe He, Xiaoxuan Mu, Lianrui Cai, Chengfei Hu, Haoji Computer Vision and Pattern Recognition Quantization has emerged as a promising direction for model compression. Recently, data-free quantization has been widely studied as a promising method to avoid privacy concerns, which synthesizes images as an alternative to real training data. Existing methods use classification loss to ensure the reliability of the synthesized images. Unfortunately, even if these images are well-classified by the pre-trained model, they still suffer from low semantics and homogenization issues. Intuitively, these low-semantic images are sensitive to perturbations, and the pre-trained model tends to have inconsistent output when the generator synthesizes an image with poor semantics. To this end, we propose Robustness-Guided Image Synthesis (RIS), a simple but effective method to enrich the semantics of synthetic images and improve image diversity, further boosting the performance of downstream data-free compression tasks. Concretely, we first introduce perturbations on input and model weight, then define the inconsistency metrics at feature and prediction levels before and after perturbations. On the basis of inconsistency on two levels, we design a robustness optimization objective to enhance the semantics of synthetic images. Moreover, we also make our approach diversity-aware by forcing the generator to synthesize images with small correlations in the label space. With RIS, we achieve state-of-the-art performance for various settings on data-free quantization and can be extended to other data-free compression tasks. |
| title | Robustness-Guided Image Synthesis for Data-Free Quantization |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2310.03661 |