Zero-Shot Low Light Image Enhancement with Diffusion Prior
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_ | 1866915210858594304 |
|---|---|
| author | Cho, Joshua Aghajanzadeh, Sara Zhu, Zhen Forsyth, D. A. |
| author_facet | Cho, Joshua Aghajanzadeh, Sara Zhu, Zhen Forsyth, D. A. |
| contents | In this paper, we present a simple yet highly effective "free lunch" solution for low-light image enhancement (LLIE), which aims to restore low-light images as if acquired in well-illuminated environments. Our method necessitates no optimization, training, fine-tuning, text conditioning, or hyperparameter adjustments, yet it consistently reconstructs low-light images with superior fidelity. Specifically, we leverage a pre-trained text-to-image diffusion prior, learned from training on a large collection of natural images, and the features present in the model itself to guide the inference, in contrast to existing methods that depend on customized constraints. Comprehensive quantitative evaluations demonstrate that our approach outperforms SOTA methods on established datasets, while qualitative analyses indicate enhanced color accuracy and the rectification of subtle chromatic deviations. Furthermore, additional experiments reveal that our method, without any modifications, achieves SOTA-comparable performance in the auto white balance (AWB) task. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2412_13401 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | Zero-Shot Low Light Image Enhancement with Diffusion Prior Cho, Joshua Aghajanzadeh, Sara Zhu, Zhen Forsyth, D. A. Computer Vision and Pattern Recognition In this paper, we present a simple yet highly effective "free lunch" solution for low-light image enhancement (LLIE), which aims to restore low-light images as if acquired in well-illuminated environments. Our method necessitates no optimization, training, fine-tuning, text conditioning, or hyperparameter adjustments, yet it consistently reconstructs low-light images with superior fidelity. Specifically, we leverage a pre-trained text-to-image diffusion prior, learned from training on a large collection of natural images, and the features present in the model itself to guide the inference, in contrast to existing methods that depend on customized constraints. Comprehensive quantitative evaluations demonstrate that our approach outperforms SOTA methods on established datasets, while qualitative analyses indicate enhanced color accuracy and the rectification of subtle chromatic deviations. Furthermore, additional experiments reveal that our method, without any modifications, achieves SOTA-comparable performance in the auto white balance (AWB) task. |
| title | Zero-Shot Low Light Image Enhancement with Diffusion Prior |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2412.13401 |