While real-world anime super-resolution (SR) has gained increasing attention in the SR community, existing methods still adopt techniques from the photorealistic domain. In this paper, we analyze the anime production workflow and rethink how to use characteristics of it for the sake of the real-world anime SR. First, we argue that video networks and datasets are not necessary for anime SR due to the repetition use of hand-drawing frames. Instead, we propose an anime image collection pipeline by choosing the least compressed and the most informative frames from the video sources. Based on this pipeline, we introduce the Anime Production-oriented Image (API) dataset. In addition, we identify two anime-specific challenges of distorted and faint hand-drawn lines and unwanted color artifacts. We address the first issue by introducing a prediction-oriented compression module in the image degradation model and a pseudo-ground truth preparation with enhanced hand-drawn lines. In addition, we introduce the balanced twin perceptual loss combining both anime and photorealistic high-level features to mitigate unwanted color artifacts and increase visual clarity. We evaluate our method through extensive experiments on the public benchmark, showing our method outperforms state-of-the-art anime dataset-trained approaches.

WorkFlow

Visual Results

Citation

@article{wang2024apisr,
  title={APISR: Anime Production Inspired Real-World Anime Super-Resolution},
  author={Wang, Boyang and Yang, Fengyu and Yu, Xihang and Zhang, Chao and Zhao, Hanbin},
  journal={arXiv preprint arXiv:2403.01598},
  year={2024}
}

Disclaimer

This dataset is provided exclusively for personal and academic research use only. Commercial use, redistribution, and creation of derivative datasets are prohibited. The dataset is provided “as-is” without any warranty or liability of any kind.

Downloads last month: 33