Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Preprint |
| Published: |
2023
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2312.00739 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866914945272119296 |
|---|---|
| author | Wei, Min Zhou, Jingkai Sun, Junyao Zhang, Xuesong |
| author_facet | Wei, Min Zhou, Jingkai Sun, Junyao Zhang, Xuesong |
| contents | Existing score distillation methods are sensitive to classifier-free guidance (CFG) scale: manifested as over-smoothness or instability at small CFG scales, while over-saturation at large ones. To explain and analyze these issues, we revisit the derivation of Score Distillation Sampling (SDS) and decipher existing score distillation with the Wasserstein Generative Adversarial Network (WGAN) paradigm. With the WGAN paradigm, we find that existing score distillation either employs a fixed sub-optimal discriminator or conducts incomplete discriminator optimization, resulting in the scale-sensitive issue. We propose the Adversarial Score Distillation (ASD), which maintains an optimizable discriminator and updates it using the complete optimization objective. Experiments show that the proposed ASD performs favorably in 2D distillation and text-to-3D tasks against existing methods. Furthermore, to explore the generalization ability of our WGAN paradigm, we extend ASD to the image editing task, which achieves competitive results. The project page and code are at https://github.com/2y7c3/ASD. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2312_00739 |
| institution | arXiv |
| publishDate | 2023 |
| record_format | arxiv |
| spellingShingle | Adversarial Score Distillation: When score distillation meets GAN Wei, Min Zhou, Jingkai Sun, Junyao Zhang, Xuesong Computer Vision and Pattern Recognition Existing score distillation methods are sensitive to classifier-free guidance (CFG) scale: manifested as over-smoothness or instability at small CFG scales, while over-saturation at large ones. To explain and analyze these issues, we revisit the derivation of Score Distillation Sampling (SDS) and decipher existing score distillation with the Wasserstein Generative Adversarial Network (WGAN) paradigm. With the WGAN paradigm, we find that existing score distillation either employs a fixed sub-optimal discriminator or conducts incomplete discriminator optimization, resulting in the scale-sensitive issue. We propose the Adversarial Score Distillation (ASD), which maintains an optimizable discriminator and updates it using the complete optimization objective. Experiments show that the proposed ASD performs favorably in 2D distillation and text-to-3D tasks against existing methods. Furthermore, to explore the generalization ability of our WGAN paradigm, we extend ASD to the image editing task, which achieves competitive results. The project page and code are at https://github.com/2y7c3/ASD. |
| title | Adversarial Score Distillation: When score distillation meets GAN |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2312.00739 |