Saved in:
| Main Authors: | , , , , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2501.08332 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866917892654628864 |
|---|---|
| author | Liu, Zhiheng Cheng, Ka Leong Chen, Xi Xiao, Jie Ouyang, Hao Zhu, Kai Liu, Yu Shen, Yujun Chen, Qifeng Luo, Ping |
| author_facet | Liu, Zhiheng Cheng, Ka Leong Chen, Xi Xiao, Jie Ouyang, Hao Zhu, Kai Liu, Yu Shen, Yujun Chen, Qifeng Luo, Ping |
| contents | Derived from diffusion models, MangaNinjia specializes in the task of reference-guided line art colorization. We incorporate two thoughtful designs to ensure precise character detail transcription, including a patch shuffling module to facilitate correspondence learning between the reference color image and the target line art, and a point-driven control scheme to enable fine-grained color matching. Experiments on a self-collected benchmark demonstrate the superiority of our model over current solutions in terms of precise colorization. We further showcase the potential of the proposed interactive point control in handling challenging cases, cross-character colorization, multi-reference harmonization, beyond the reach of existing algorithms. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2501_08332 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | MangaNinja: Line Art Colorization with Precise Reference Following Liu, Zhiheng Cheng, Ka Leong Chen, Xi Xiao, Jie Ouyang, Hao Zhu, Kai Liu, Yu Shen, Yujun Chen, Qifeng Luo, Ping Computer Vision and Pattern Recognition Derived from diffusion models, MangaNinjia specializes in the task of reference-guided line art colorization. We incorporate two thoughtful designs to ensure precise character detail transcription, including a patch shuffling module to facilitate correspondence learning between the reference color image and the target line art, and a point-driven control scheme to enable fine-grained color matching. Experiments on a self-collected benchmark demonstrate the superiority of our model over current solutions in terms of precise colorization. We further showcase the potential of the proposed interactive point control in handling challenging cases, cross-character colorization, multi-reference harmonization, beyond the reach of existing algorithms. |
| title | MangaNinja: Line Art Colorization with Precise Reference Following |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2501.08332 |