{"url":"/sota/3d-semantic-segmentation-on-scannet200","task":{"name":"3D Semantic Segmentation","url":"/task/3d-semantic-segmentation","note":null},"dataset":{"name":"ScanNet200","url":"/dataset/scannet200"},"category":"Computer Vision","categories":["Computer Vision"],"category_note":null,"description":"**3D Semantic Segmentation** is a computer vision task that involves dividing a 3D point cloud or 3D mesh into semantically meaningful parts or regions. The goal of 3D semantic segmentation is to identify and label different objects and parts within a 3D scene, which can be used for applications such as robotics, autonomous driving, and augmented reality.","description_from":"task","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","rank":"the archive's row order at snapshot; not re-ranked","rows_end_at":"2025-07-28","rows_withheld_as_spam":0,"metric_values":"the archive's strings, untouched"},"metrics":["val mIoU","test mIoU"],"metric_direction":{"note":"inferred from the metric name only (the archive records no direction); null = not inferred, chart draws points only","by_metric":{"val mIoU":null,"test mIoU":null}},"counts":{"rows":16,"rows_with_code":15,"rows_with_paper_page":16,"rows_dated":16,"rows_using_additional_data":5},"rows":[{"rank_in_archive_order":1,"model":"DITR","metrics":{"test mIoU":"44.9","val mIoU":"41.2"},"uses_additional_data":false,"paper_date":"2025-03-24","paper":"/paper/dino-in-the-room-leveraging-2d-foundation","paper_url":"https://arxiv.org/abs/2503.18944v1","paper_title":"DINO in the Room: Leveraging 2D Foundation Models for 3D Segmentation","code":"https://github.com/VisualComputingInstitute/DITR","n_code_links":1,"syntology":null},{"rank_in_archive_order":2,"model":"ODIN","metrics":{"test mIoU":"36.8","val mIoU":"40.5"},"uses_additional_data":false,"paper_date":"2024-01-04","paper":"/paper/odin-a-single-model-for-2d-and-3d-perception","paper_url":"https://arxiv.org/abs/2401.02416v3","paper_title":"ODIN: A Single Model for 2D and 3D Segmentation","code":"https://github.com/ayushjain1144/odin","n_code_links":1,"syntology":{"n_ran":3,"n_unverified":5,"n_samples":8,"n_pointer_only_licence":0}},{"rank_in_archive_order":3,"model":"PTv3 ArKitLabelmaker","metrics":{"test mIoU":"41.4","val mIoU":"40.3"},"uses_additional_data":true,"paper_date":"2024-10-17","paper":"/paper/arkit-labelmaker-a-new-scale-for-indoor-3d","paper_url":"https://arxiv.org/abs/2410.13924v1","paper_title":"ARKit LabelMaker: A New Scale for Indoor 3D Scene Understanding","code":"https://github.com/cvg/labelmaker","n_code_links":1,"syntology":{"n_ran":2,"n_unverified":0,"n_samples":2,"n_pointer_only_licence":2}},{"rank_in_archive_order":4,"model":"BFANet","metrics":{"test mIoU":"36.0","val mIoU":"37.3"},"uses_additional_data":false,"paper_date":"2025-03-16","paper":"/paper/bfanet-revisiting-3d-semantic-segmentation","paper_url":"https://arxiv.org/abs/2503.12539v1","paper_title":"BFANet: Revisiting 3D Semantic Segmentation with Boundary Feature Analysis","code":"https://github.com/weiguangzhao/BFANet","n_code_links":1,"syntology":null},{"rank_in_archive_order":5,"model":"Sonata + PTv3","metrics":{"val mIoU":"36.8"},"uses_additional_data":true,"paper_date":"2025-03-20","paper":"/paper/sonata-self-supervised-learning-of-reliable","paper_url":"https://arxiv.org/abs/2503.16429v1","paper_title":"Sonata: Self-Supervised Learning of Reliable Point Representations","code":"https://github.com/Pointcept/Pointcept","n_code_links":2,"syntology":{"n_ran":10,"n_unverified":9,"n_samples":19,"n_pointer_only_licence":0}},{"rank_in_archive_order":6,"model":"Pamba","metrics":{"test mIoU":"37.1","val mIoU":"36.3"},"uses_additional_data":false,"paper_date":"2024-06-25","paper":"/paper/mamba24-8d-enhancing-global-interaction-in","paper_url":"https://arxiv.org/abs/2406.17442v3","paper_title":"Pamba: Enhancing Global Interaction in Point Clouds via State Space Model","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":7,"model":"PTv3 + PPT","metrics":{"test mIoU":"39.3","val mIoU":"36.0"},"uses_additional_data":true,"paper_date":"2023-12-15","paper":"/paper/point-transformer-v3-simpler-faster-stronger","paper_url":"https://arxiv.org/abs/2312.10035v2","paper_title":"Point Transformer V3: Simpler, Faster, Stronger","code":"https://github.com/Pointcept/Pointcept","n_code_links":3,"syntology":{"n_ran":7,"n_unverified":4,"n_samples":11,"n_pointer_only_licence":0}},{"rank_in_archive_order":8,"model":"OA-CNNs","metrics":{"test mIoU":"32.3","val mIoU":"33.3"},"uses_additional_data":false,"paper_date":"2024-03-21","paper":"/paper/oa-cnns-omni-adaptive-sparse-cnns-for-3d","paper_url":"https://arxiv.org/abs/2403.14418v1","paper_title":"OA-CNNs: Omni-Adaptive Sparse CNNs for 3D Semantic Segmentation","code":"https://github.com/Pointcept/Pointcept","n_code_links":1,"syntology":null},{"rank_in_archive_order":9,"model":"LSK3DNet","metrics":{"val mIoU":"33.1"},"uses_additional_data":false,"paper_date":"2024-03-22","paper":"/paper/lsk3dnet-towards-effective-and-efficient-3d","paper_url":"https://arxiv.org/abs/2403.15173v1","paper_title":"LSK3DNet: Towards Effective and Efficient 3D Perception with Large Sparse Kernels","code":"https://github.com/fengzicai/lsk3dnet","n_code_links":1,"syntology":null},{"rank_in_archive_order":10,"model":"OctFormer","metrics":{"test mIoU":"32.5","val mIoU":"32.6"},"uses_additional_data":false,"paper_date":"2023-05-04","paper":"/paper/octformer-octree-based-transformers-for-3d","paper_url":"https://arxiv.org/abs/2305.03045v2","paper_title":"OctFormer: Octree-based Transformers for 3D Point Clouds","code":"https://github.com/Pointcept/Pointcept","n_code_links":4,"syntology":{"n_ran":6,"n_unverified":3,"n_samples":9,"n_pointer_only_licence":0}},{"rank_in_archive_order":11,"model":"PonderV2 + SparseUNet","metrics":{"test mIoU":"34.6","val mIoU":"32.3"},"uses_additional_data":true,"paper_date":"2023-10-12","paper":"/paper/ponderv2-pave-the-way-for-3d-foundataion","paper_url":"https://arxiv.org/abs/2310.08586v3","paper_title":"PonderV2: Pave the Way for 3D Foundation Model with A Universal Pre-training Paradigm","code":"https://github.com/OpenGVLab/PonderV2","n_code_links":1,"syntology":null},{"rank_in_archive_order":12,"model":"PPT+SparseUNet","metrics":{"test mIoU":"33.2","val mIoU":"31.9"},"uses_additional_data":true,"paper_date":"2023-08-18","paper":"/paper/towards-large-scale-3d-representation","paper_url":"https://arxiv.org/abs/2308.09718v2","paper_title":"Towards Large-scale 3D Representation Learning with Multi-dataset Point Prompt Training","code":"https://github.com/Pointcept/Pointcept","n_code_links":1,"syntology":null},{"rank_in_archive_order":13,"model":"OneFormer3D","metrics":{"val mIoU":"30.1"},"uses_additional_data":false,"paper_date":"2023-11-24","paper":"/paper/oneformer3d-one-transformer-for-unified-point","paper_url":"https://arxiv.org/abs/2311.14405v1","paper_title":"OneFormer3D: One Transformer for Unified Point Cloud Segmentation","code":"https://github.com/oneformer3d/oneformer3d","n_code_links":1,"syntology":null},{"rank_in_archive_order":14,"model":"LGround","metrics":{"test mIoU":"27.2","val mIoU":"28.8"},"uses_additional_data":false,"paper_date":"2022-04-16","paper":"/paper/language-grounded-indoor-3d-semantic","paper_url":"https://arxiv.org/abs/2204.07761v2","paper_title":"Language-Grounded Indoor 3D Semantic Segmentation in the Wild","code":"https://github.com/RozDavid/LanguageGroundedSemseg","n_code_links":1,"syntology":{"n_ran":2,"n_unverified":0,"n_samples":2,"n_pointer_only_licence":2}},{"rank_in_archive_order":15,"model":"CSC","metrics":{"test mIoU":"24.9","val mIoU":"26.4"},"uses_additional_data":false,"paper_date":"2020-12-16","paper":"/paper/exploring-data-efficient-3d-scene","paper_url":"https://arxiv.org/abs/2012.09165v3","paper_title":"Exploring Data-Efficient 3D Scene Understanding with Contrastive Scene Contexts","code":"https://github.com/Pointcept/Pointcept","n_code_links":2,"syntology":null},{"rank_in_archive_order":16,"model":"MinkUNet","metrics":{"test mIoU":"25.3","val mIoU":"25.0"},"uses_additional_data":false,"paper_date":"2019-04-18","paper":"/paper/4d-spatio-temporal-convnets-minkowski","paper_url":"https://arxiv.org/abs/1904.08755v4","paper_title":"4D Spatio-Temporal ConvNets: Minkowski Convolutional Neural Networks","code":"https://github.com/NVIDIA/MinkowskiEngine","n_code_links":8,"syntology":{"n_ran":1,"n_unverified":1,"n_samples":2,"n_pointer_only_licence":2}}],"since_archive":{"claim":"Results that newer papers report for their own method, placed here by Syntology. A model pointed at the cell in the paper's own table; the number was read from that cell and checked against this leaderboard's metric, dataset, split and scale; an independent check that saw this leaderboard's other rows and every other leaderboard on the same dataset accepted it. Not reviewed by the paper's authors or by the archive's editors, and not ranked against the archive rows.","extraction_file_present":true,"measurement":{"test_papers":883,"papers_with_output":881,"judged_true":108,"judged":110,"wilson95_lower":0.9361,"measured_on":"2026-09-24","frozen_commit":"0e3de0df94"},"measurement_note":"blind adjudication of accepted entries on a held-out split of archive papers, rules frozen before the test","coverage":{"sentence":"Syntology has checked 6,885 of the 9,623 papers on this site that are newer than the archive; results from the others appear after they are checked.","complete":false,"papers_newer_than_archive":9623,"papers_checked":6885,"papers_extracted_not_yet_verified":0,"boards_without_verdict":2,"papers_not_yet_extracted":2737},"order":"newest first by month (arXiv date, else the arXiv-id month), then arXiv id descending","columns":[],"entries":[]},"syntology":{"read_at":"2026-09-25T09:33:49+00:00","claim":"Per row: N of M harvested code samples from that row's paper executed on a synthesized fixture; the other M-N are unverified. Not a reproduction of the row's number; not a correctness claim. n_pointer_only_licence counts samples the site points at rather than redistributes (a licence axis, independent of ran/unverified).","rows_with_graph_line":7,"rows_with_any_sample_ran":7,"distinct_papers_with_graph_line":7,"distinct_papers_with_any_sample_ran":7,"samples_over_distinct_papers":{"n_ran":31,"n_unverified":22,"n_samples":53,"n_pointer_only_licence":6,"note":"each paper (arXiv id) counted once, however many rows it is behind; this is the page-level figure"},"samples_row_weighted":{"n_ran":31,"n_unverified":22,"n_samples":53,"n_pointer_only_licence":6,"note":"row-weighted: a paper behind several rows is counted once per row; inflated relative to samples_over_distinct_papers by design, kept for readers summing the per-row syntology blocks"}}}