{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/blazingly-fast-video-object-segmentation-with","title":"Blazingly Fast Video Object Segmentation with Pixel-Wise Metric Learning","arxiv_id":"1804.03131","date":"2018-04-09","proceeding":"CVPR 2018 6","authors":["Yuhua Chen","Jordi Pont-Tuset","Alberto Montes","Luc van Gool"],"abstract":"This paper tackles the problem of video object segmentation, given some user\nannotation which indicates the object of interest. The problem is formulated as\npixel-wise retrieval in a learned embedding space: we embed pixels of the same\nobject instance into the vicinity of each other, using a fully convolutional\nnetwork trained by a modified triplet loss as the embedding model. Then the\nannotated pixels are set as reference and the rest of the pixels are classified\nusing a nearest-neighbor approach. The proposed method supports different kinds\nof user input such as segmentation mask in the first frame (semi-supervised\nscenario), or a sparse set of clicked points (interactive scenario). In the\nsemi-supervised scenario, we achieve results competitive with the state of the\nart but at a fraction of computation cost (275 milliseconds per frame). In the\ninteractive scenario where the user is able to refine their input iteratively,\nthe proposed method provides instant response to each input, and reaches\ncomparable quality to competing methods with much less interaction.","url_abs":"http://arxiv.org/abs/1804.03131v1","url_pdf":"http://arxiv.org/pdf/1804.03131v1.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[],"tasks":[{"task_slug":"metric-learning","task_name":"Metric Learning"},{"task_slug":"object","task_name":"Object"},{"task_slug":"retrieval","task_name":"Retrieval"},{"task_slug":"segmentation","task_name":"Segmentation"},{"task_slug":"semantic-segmentation","task_name":"Semantic Segmentation"},{"task_slug":"semi-supervised-video-object-segmentation","task_name":"Semi-Supervised Video Object Segmentation"},{"task_slug":null,"task_name":"Triplet"},{"task_slug":"video-object-segmentation","task_name":"Video Object Segmentation"},{"task_slug":"video-semantic-segmentation","task_name":"Video Semantic Segmentation"},{"task_slug":"visual-object-tracking","task_name":"Visual Object Tracking"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/visual-object-tracking-on-davis-2016","task":"Semi-Supervised Video Object Segmentation","dataset":"DAVIS 2016","model":"PML","rank_in_archive_order":66,"of":78,"metrics":{"F-measure (Decay)":"7.8","F-measure (Mean)":"79.3","F-measure (Recall)":"93.4","J&F":"77.4","Jaccard (Decay)":"8.5","Jaccard (Mean)":"75.5","Jaccard (Recall)":"89.6"},"uses_additional_data":false}],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1804.03131","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}