{"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/masc-multi-scale-affinity-with-sparse","title":"MASC: Multi-scale Affinity with Sparse Convolution for 3D Instance Segmentation","arxiv_id":"1902.04478","date":"2019-02-12","proceeding":null,"authors":["Chen Liu","Yasutaka Furukawa"],"abstract":"We propose a new approach for 3D instance segmentation based on sparse\nconvolution and point affinity prediction, which indicates the likelihood of\ntwo points belonging to the same instance. The proposed network, built upon\nsubmanifold sparse convolution [3], processes a voxelized point cloud and\npredicts semantic scores for each occupied voxel as well as the affinity\nbetween neighboring voxels at different scales. A simple yet effective\nclustering algorithm segments points into instances based on the predicted\naffinity and the mesh topology. The semantic for each instance is determined by\nthe semantic prediction. Experiments show that our method outperforms the\nstate-of-the-art instance segmentation methods by a large margin on the widely\nused ScanNet benchmark [2]. We share our code publicly at\nhttps://github.com/art-programmer/MASC.","url_abs":"http://arxiv.org/abs/1902.04478v1","url_pdf":"http://arxiv.org/pdf/1902.04478v1.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":[{"paper_slug":"masc-multi-scale-affinity-with-sparse","repo_url":"https://github.com/art-programmer/MASC","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"pytorch","reach":null}],"tasks":[{"task_slug":"3d-instance-segmentation-1","task_name":"3D Instance Segmentation"},{"task_slug":"clustering","task_name":"Clustering"},{"task_slug":"instance-segmentation","task_name":"Instance Segmentation"},{"task_slug":"segmentation","task_name":"Segmentation"},{"task_slug":"semantic-segmentation","task_name":"Semantic Segmentation"}],"methods":[{"method_slug":"convolution","method_name":"Convolution"}],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/3d-instance-segmentation-on-scannet","task":"3D Instance Segmentation","dataset":"ScanNet","model":"MASC","rank_in_archive_order":1,"of":1,"metrics":{"mAP":"0.447"},"uses_additional_data":false},{"leaderboard":"/sota/3d-instance-segmentation-on-scannetv2","task":"3D Instance Segmentation","dataset":"ScanNet(v2)","model":"ResNet-Backbone","rank_in_archive_order":26,"of":32,"metrics":{"mAP @ 50":"45.9"},"uses_additional_data":false},{"leaderboard":"/sota/3d-instance-segmentation-on-scannetv2","task":"3D Instance Segmentation","dataset":"ScanNet(v2)","model":"MASC","rank_in_archive_order":27,"of":32,"metrics":{"mAP":"25.4","mAP @ 50":"44.7"},"uses_additional_data":false}],"syntology":{"syntology_url":"https://syntology.ai/paper/1902.04478","atlas_url":"https://app.syntology.ai/?focus=1902.04478","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}