{"url":"/method/modnet","slug":"modnet","name":"MODNet","full_name":"MODNet","full_name_withheld":false,"description_markdown":"**MODNet** is a light-weight matting objective decomposition network that can process portrait matting from a single input image in real time. The design of MODNet benefits from optimizing a series of correlated sub-objectives simultaneously via explicit constraints. To overcome the domain shift problem, MODNet introduces a self-supervised strategy based on subobjective consistency (SOC) and  a one-frame delay trick to smooth the results when applying MODNet to portrait video sequence.\r\n\r\nGiven an input image $I$, MODNet predicts human semantics $s\\_{p}$, boundary details $d\\_{p}$, and final alpha matte $\\alpha\\_{p}$ through three interdependent branches, $S, D$, and $F$, which are constrained by specific supervisions generated from the ground truth matte $\\alpha\\_{g}$. Since the decomposed sub-objectives are correlated and help strengthen each other, we can optimize MODNet end-to-end.","description_state":"present","introduced_year":null,"introduced_by":{"title":"MODNet: Real-Time Trimap-Free Portrait Matting via Objective Decomposition","paper":"/paper/is-a-green-screen-really-necessary-for-real","first_author":"Zhanghan Ke","n_authors":5,"url_abs":null,"archive_paper_url":"https://paperswithcode.com/paper/is-a-green-screen-really-necessary-for-real"},"source":{"url":"https://arxiv.org/abs/2011.11961v4","title":"MODNet: Real-Time Trimap-Free Portrait Matting via Objective Decomposition","url_on_a_paper_host":true},"code_snippet_url":null,"code_snippet_url_on_a_code_host":false,"categories":[{"area":"Computer Vision","area_id":"computer-vision","collection":"Portrait Matting Models","url":"/methods/category/portrait-matting-models","pwc_aliases":[]}],"n_papers_tagged":4,"archive_num_papers":4,"papers_newest_first":[{"paper":null,"title":"SGM-Net: Semantic Guided Matting Net","date":"2022-08-16","arxiv_id":"2208.07496","n_code_links":0,"syntology":null},{"paper":"/paper/modnet-v-improving-portrait-video-matting-via","title":"MODNet-V: Improving Portrait Video Matting via Background Restoration","date":"2021-09-24","arxiv_id":"2109.11818","n_code_links":1,"syntology":null},{"paper":null,"title":"Alpha Matte Generation from Single Input for Portrait Matting","date":"2021-06-06","arxiv_id":"2106.03210","n_code_links":0,"syntology":null},{"paper":"/paper/is-a-green-screen-really-necessary-for-real","title":"MODNet: Real-Time Trimap-Free Portrait Matting via Objective Decomposition","date":"2020-11-24","arxiv_id":"2011.11961","n_code_links":9,"syntology":null}],"papers_shown":4,"tasks":[{"task":"/task/image-matting","name":"Image Matting","papers":4},{"task":null,"name":"GPU","papers":2},{"task":"/task/video-matting","name":"Video Matting","papers":2},{"task":"/task/image-generation","name":"Image Generation","papers":1},{"task":"/task/segmentation","name":"Segmentation","papers":1}],"tasks_shown":5,"n_tasks":5,"usage_by_year":[{"year":"2020","papers":1},{"year":"2021","papers":2},{"year":"2022","papers":1}],"row_source":"methods_table","archive":{"source":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","archive_url":"https://paperswithcode.com/method/modnet"},"syntology_read_at":"2026-09-24T18:15:14+00:00"}