{"ID":2873522,"CreatedAt":"2026-06-01T04:54:23.091178241Z","UpdatedAt":"2026-06-01T04:54:23.091178241Z","DeletedAt":null,"paper_url":"https://arxiv.org/abs/2509.06839","arxiv_id":"2509.06839","title":"ToonOut: Fine-tuned Background-Removal for Anime Characters","abstract":"While state-of-the-art background removal models excel at realistic imagery, they frequently underperform in specialized domains such as anime-style content, where complex features like hair and transparency present unique challenges. To address this limitation, we collected and annotated a custom dataset of 1,228 high-quality anime images of characters and objects, and fine-tuned the open-sourced BiRefNet model on this dataset. This resulted in marked improvements in background removal accuracy for anime-style images, increasing from 95.3% to 99.5% for our newly introduced Pixel Accuracy metric. We are open-sourcing the code, the fine-tuned model weights, as well as the dataset at: https://github.com/MatteoKartoon/BiRefNet.","short_abstract":"While state-of-the-art background removal models excel at realistic imagery, they frequently underperform in specialized domains such as anime-style content, where complex features like hair and transparency present unique challenges. To address this limitation, we collected and annotated a custom dataset of 1,228 high...","url_abs":"https://arxiv.org/abs/2509.06839","url_pdf":"https://arxiv.org/pdf/2509.06839v1","authors":"[\"Matteo Muratori\",\"Joël Seytre\"]","published":"2025-09-08T16:08:56Z","proceeding":"cs.CV","tasks":"[\"cs.CV\",\"cs.LG\"]","methods":"[]","has_code":false,"code_links":[{"ID":610063,"CreatedAt":"2026-06-01T04:54:23.091178241Z","UpdatedAt":"2026-06-01T04:54:23.091178241Z","DeletedAt":null,"paper_id":2873522,"paper_url":"https://arxiv.org/abs/2509.06839","paper_title":"ToonOut: Fine-tuned Background-Removal for Anime Characters","repo_url":"https://github.com/MatteoKartoon/BiRefNet","is_official":false,"mentioned_in_paper":false,"mentioned_in_github":true,"github_stars":0}]}