{"url":"/sota/3d-reconstruction-on-dtu","task":{"name":"3D Reconstruction","url":"/task/3d-reconstruction","note":null},"dataset":{"name":"DTU","url":"/dataset/dtu"},"category":"Computer Vision","categories":["Computer Vision","Methodology"],"category_note":null,"description":"**3D Reconstruction** is the task of creating a 3D model or representation of an object or scene from 2D images or other data sources. The goal of 3D reconstruction is to create a virtual representation of an object or scene that can be used for a variety of purposes, such as visualization, animation, simulation, and analysis. It can be used in fields such as computer vision, robotics, and virtual reality.\r\n\r\nImage: [Gwak et al](https://arxiv.org/pdf/1705.10904v2.pdf)","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":["Overall","Acc","Comp"],"metric_direction":{"note":"inferred from the metric name only (the archive records no direction); null = not inferred, chart draws points only","by_metric":{"Overall":null,"Acc":"higher","Comp":null}},"counts":{"rows":24,"rows_with_code":22,"rows_with_paper_page":24,"rows_dated":24,"rows_using_additional_data":0},"rows":[{"rank_in_archive_order":1,"model":"MVSFormer++","metrics":{"Acc":"0.3090","Comp":"0.2521","Overall":"0.2805"},"uses_additional_data":false,"paper_date":"2024-01-22","paper":"/paper/mvsformer-revealing-the-devil-in-transformer","paper_url":"https://arxiv.org/abs/2401.11673v1","paper_title":"MVSFormer++: Revealing the Devil in Transformer's Details for Multi-View Stereo","code":"https://github.com/maybelx/mvsformerplusplus","n_code_links":1,"syntology":{"n_ran":15,"n_unverified":4,"n_samples":19,"n_pointer_only_licence":0}},{"rank_in_archive_order":2,"model":"GoMVS","metrics":{"Acc":"0.347","Comp":"0.227","Overall":"0.287"},"uses_additional_data":false,"paper_date":"2024-04-11","paper":"/paper/gomvs-geometrically-consistent-cost","paper_url":"https://arxiv.org/abs/2404.07992v1","paper_title":"GoMVS: Geometrically Consistent Cost Aggregation for Multi-View Stereo","code":"https://github.com/wuuu3511/gomvs","n_code_links":1,"syntology":{"n_ran":0,"n_unverified":6,"n_samples":6,"n_pointer_only_licence":0}},{"rank_in_archive_order":3,"model":"MVSFormer","metrics":{"Acc":"0.327","Comp":"0.251","Overall":"0.289"},"uses_additional_data":false,"paper_date":"2022-08-04","paper":"/paper/mvsformer-learning-robust-image","paper_url":"https://arxiv.org/abs/2208.02541v3","paper_title":"MVSFormer: Multi-View Stereo by Learning Robust Image Features and Temperature-based Depth","code":"https://github.com/ewrfcas/mvsformer","n_code_links":1,"syntology":{"n_ran":13,"n_unverified":2,"n_samples":15,"n_pointer_only_licence":0}},{"rank_in_archive_order":4,"model":"ET-MVSNet","metrics":{"Acc":"0.329","Comp":"0.253","Overall":"0.291"},"uses_additional_data":false,"paper_date":"2023-09-29","paper":"/paper/when-epipolar-constraint-meets-non-local-1","paper_url":"https://arxiv.org/abs/2309.17218v1","paper_title":"When Epipolar Constraint Meets Non-local Operators in Multi-View Stereo","code":"https://github.com/tqtqliu/et-mvsnet","n_code_links":1,"syntology":{"n_ran":5,"n_unverified":8,"n_samples":13,"n_pointer_only_licence":0}},{"rank_in_archive_order":5,"model":"GC-MVSNet","metrics":{"Acc":"0.330","Comp":"0.260","Overall":"0.295"},"uses_additional_data":false,"paper_date":"2023-10-30","paper":"/paper/gc-mvsnet-multi-view-multi-scale","paper_url":"https://arxiv.org/abs/2310.19583v3","paper_title":"GC-MVSNet: Multi-View, Multi-Scale, Geometrically-Consistent Multi-View Stereo","code":"https://github.com/vkvats/GC-MVSNet","n_code_links":1,"syntology":null},{"rank_in_archive_order":6,"model":"GeoMVSNet","metrics":{"Acc":"0.331","Comp":"0.259","Overall":"0.295"},"uses_additional_data":false,"paper_date":"2023-01-01","paper":"/paper/geomvsnet-learning-multi-view-stereo-with","paper_url":"http://openaccess.thecvf.com//content/CVPR2023/html/Zhang_GeoMVSNet_Learning_Multi-View_Stereo_With_Geometry_Perception_CVPR_2023_paper.html","paper_title":"GeoMVSNet: Learning Multi-View Stereo With Geometry Perception","code":"https://github.com/doublez0108/geomvsnet","n_code_links":1,"syntology":null},{"rank_in_archive_order":7,"model":"RA-MVSNet","metrics":{"Acc":"0.326","Comp":"0.268","Overall":"0.297"},"uses_additional_data":false,"paper_date":"2023-04-26","paper":"/paper/multi-view-stereo-representation-revist","paper_url":"https://arxiv.org/abs/2304.13614v2","paper_title":"Multi-View Stereo Representation Revisit: Region-Aware MVSNet","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":8,"model":"GBi-Net","metrics":{"Acc":"0.312","Comp":"0.293","Overall":"0.303"},"uses_additional_data":false,"paper_date":"2021-12-04","paper":"/paper/generalized-binary-search-network-for-highly","paper_url":"https://arxiv.org/abs/2112.02338v1","paper_title":"Generalized Binary Search Network for Highly-Efficient Multi-View Stereo","code":"https://github.com/mizhenxing/gbi-net","n_code_links":1,"syntology":null},{"rank_in_archive_order":9,"model":"TransMVSNet","metrics":{"Acc":"0.321","Comp":"0.289","Overall":"0.305"},"uses_additional_data":false,"paper_date":"2021-11-29","paper":"/paper/transmvsnet-global-context-aware-multi-view","paper_url":"https://arxiv.org/abs/2111.14600v1","paper_title":"TransMVSNet: Global Context-aware Multi-view Stereo Network with Transformers","code":"https://github.com/megviirobot/transmvsnet","n_code_links":1,"syntology":null},{"rank_in_archive_order":10,"model":"CDS-MVSNet","metrics":{"Acc":"0.351","Comp":"0.278","Overall":"0.315"},"uses_additional_data":false,"paper_date":"2021-12-11","paper":"/paper/curvature-guided-dynamic-scale-networks-for-1","paper_url":"https://arxiv.org/abs/2112.05999v3","paper_title":"Curvature-guided dynamic scale networks for Multi-view Stereo","code":"https://github.com/truongkhang/cds-mvsnet","n_code_links":1,"syntology":{"n_ran":16,"n_unverified":2,"n_samples":18,"n_pointer_only_licence":18}},{"rank_in_archive_order":11,"model":"UniMVSNet","metrics":{"Acc":"0.352","Comp":"0.278","Overall":"0.315"},"uses_additional_data":false,"paper_date":"2022-01-05","paper":"/paper/rethinking-depth-estimation-for-multi-view","paper_url":"https://arxiv.org/abs/2201.01501v3","paper_title":"Rethinking Depth Estimation for Multi-View Stereo: A Unified Representation","code":"https://github.com/prstrive/unimvsnet","n_code_links":1,"syntology":{"n_ran":9,"n_unverified":6,"n_samples":15,"n_pointer_only_licence":0}},{"rank_in_archive_order":12,"model":"IB-MVS","metrics":{"Acc":"0.334","Comp":"0.309","Overall":"0.321"},"uses_additional_data":false,"paper_date":"2021-11-29","paper":"/paper/ib-mvs-an-iterative-algorithm-for-deep-multi","paper_url":"https://arxiv.org/abs/2111.14420v1","paper_title":"IB-MVS: An Iterative Algorithm for Deep Multi-View Stereo based on Binary Decisions","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":13,"model":"MSCVP-MVSNet","metrics":{"Acc":"0.379","Comp":"0.278","Overall":"0.328"},"uses_additional_data":false,"paper_date":"2022-07-25","paper":"/paper/cost-volume-pyramid-network-with-multi","paper_url":"https://arxiv.org/abs/2207.12032v1","paper_title":"Cost Volume Pyramid Network with Multi-strategies Range Searching for Multi-view Stereo","code":"https://github.com/SibylGao/MSCVP-MVSNet","n_code_links":1,"syntology":null},{"rank_in_archive_order":14,"model":"UCSNet","metrics":{"Acc":"0.338","Comp":"0.349","Overall":"0.344"},"uses_additional_data":false,"paper_date":"2019-11-27","paper":"/paper/deep-stereo-using-adaptive-thin-volume","paper_url":"https://arxiv.org/abs/1911.12012v2","paper_title":"Deep Stereo using Adaptive Thin Volume Representation with Uncertainty Awareness","code":"https://github.com/touristCheng/UCSNet","n_code_links":1,"syntology":{"n_ran":0,"n_unverified":1,"n_samples":1,"n_pointer_only_licence":0}},{"rank_in_archive_order":15,"model":"CVP-MVSNet","metrics":{"Acc":"0.296","Comp":"0.406","Overall":"0.351"},"uses_additional_data":false,"paper_date":"2019-12-18","paper":"/paper/cost-volume-pyramid-based-depth-inference-for","paper_url":"https://arxiv.org/abs/1912.08329v3","paper_title":"Cost Volume Pyramid Based Depth Inference for Multi-View Stereo","code":"https://github.com/JiayuYANG/CVP-MVSNet","n_code_links":1,"syntology":null},{"rank_in_archive_order":16,"model":"PatchmatchNet","metrics":{"Acc":"0.427","Comp":"0.277","Overall":"0.352"},"uses_additional_data":false,"paper_date":"2020-12-02","paper":"/paper/patchmatchnet-learned-multi-view-patchmatch","paper_url":"https://arxiv.org/abs/2012.01411v1","paper_title":"PatchmatchNet: Learned Multi-View Patchmatch Stereo","code":"https://github.com/FangjinhuaWang/PatchmatchNet","n_code_links":1,"syntology":null},{"rank_in_archive_order":17,"model":"Cas-MVSNet","metrics":{"Acc":"0.325","Comp":"0.385","Overall":"0.355"},"uses_additional_data":false,"paper_date":"2019-12-13","paper":"/paper/cascade-cost-volume-for-high-resolution-multi","paper_url":"https://arxiv.org/abs/1912.06378v3","paper_title":"Cascade Cost Volume for High-Resolution Multi-View Stereo and Stereo Matching","code":"https://github.com/apchenstu/mvsnerf","n_code_links":4,"syntology":{"n_ran":6,"n_unverified":8,"n_samples":14,"n_pointer_only_licence":0}},{"rank_in_archive_order":18,"model":"EPP-MVSNet","metrics":{"Acc":"0.413","Comp":"0.296","Overall":"0.355"},"uses_additional_data":false,"paper_date":"2021-01-01","paper":"/paper/epp-mvsnet-epipolar-assembling-based-depth","paper_url":"http://openaccess.thecvf.com//content/ICCV2021/html/Ma_EPP-MVSNet_Epipolar-Assembling_Based_Depth_Prediction_for_Multi-View_Stereo_ICCV_2021_paper.html","paper_title":"EPP-MVSNet: Epipolar-Assembling Based Depth Prediction for Multi-View Stereo","code":"https://github.com/mindspore-ai/models/tree/master/research/cv/eppmvsnet","n_code_links":6,"syntology":null},{"rank_in_archive_order":19,"model":"AA-RMVSNet","metrics":{"Acc":"0.376","Comp":"0.339","Overall":"0.357"},"uses_additional_data":false,"paper_date":"2021-08-09","paper":"/paper/aa-rmvsnet-adaptive-aggregation-recurrent","paper_url":"https://arxiv.org/abs/2108.03824v1","paper_title":"AA-RMVSNet: Adaptive Aggregation Recurrent Multi-view Stereo Network","code":"https://github.com/qt-zhu/aa-rmvsnet","n_code_links":1,"syntology":{"n_ran":4,"n_unverified":16,"n_samples":20,"n_pointer_only_licence":0}},{"rank_in_archive_order":20,"model":"Vis-MVSNet","metrics":{"Acc":"0.369","Comp":"0.361","Overall":"0.365"},"uses_additional_data":false,"paper_date":"2020-08-18","paper":"/paper/visibility-aware-multi-view-stereo-network","paper_url":"https://arxiv.org/abs/2008.07928v2","paper_title":"Visibility-aware Multi-view Stereo Network","code":"https://github.com/jzhangbs/Vis-MVSNet","n_code_links":1,"syntology":{"n_ran":3,"n_unverified":11,"n_samples":14,"n_pointer_only_licence":0}},{"rank_in_archive_order":21,"model":"MVSNet","metrics":{"Acc":"0.396","Comp":"0.527","Overall":"0.462"},"uses_additional_data":false,"paper_date":"2018-04-07","paper":"/paper/mvsnet-depth-inference-for-unstructured-multi","paper_url":"http://arxiv.org/abs/1804.02505v2","paper_title":"MVSNet: Depth Inference for Unstructured Multi-view Stereo","code":"https://github.com/YoYo000/MVSNet","n_code_links":5,"syntology":null},{"rank_in_archive_order":22,"model":"COLMAP","metrics":{"Acc":"0.400","Comp":"0.664","Overall":"0.532"},"uses_additional_data":false,"paper_date":"2016-06-01","paper":"/paper/structure-from-motion-revisited","paper_url":"http://openaccess.thecvf.com/content_cvpr_2016/html/Schonberger_Structure-From-Motion_Revisited_CVPR_2016_paper.html","paper_title":"Structure-From-Motion Revisited","code":"https://github.com/colmap/colmap","n_code_links":1,"syntology":null},{"rank_in_archive_order":23,"model":"Gipuma","metrics":{"Acc":"0.283","Comp":"0.873","Overall":"0.578"},"uses_additional_data":false,"paper_date":"2015-12-01","paper":"/paper/massively-parallel-multiview-stereopsis-by","paper_url":"http://openaccess.thecvf.com/content_iccv_2015/html/Galliani_Massively_Parallel_Multiview_ICCV_2015_paper.html","paper_title":"Massively Parallel Multiview Stereopsis by Surface Normal Diffusion","code":"https://github.com/kysucix/gipuma","n_code_links":1,"syntology":null},{"rank_in_archive_order":24,"model":"3D-R2N2","metrics":{"Acc":"0.397","Comp":"0.884","Overall":"0.630"},"uses_additional_data":false,"paper_date":"2016-04-02","paper":"/paper/3d-r2n2-a-unified-approach-for-single-and","paper_url":"http://arxiv.org/abs/1604.00449v1","paper_title":"3D-R2N2: A Unified Approach for Single and Multi-view 3D Object Reconstruction","code":"https://github.com/chrischoy/3D-R2N2","n_code_links":13,"syntology":null}],"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":10,"rows_with_any_sample_ran":8,"distinct_papers_with_graph_line":10,"distinct_papers_with_any_sample_ran":8,"samples_over_distinct_papers":{"n_ran":71,"n_unverified":64,"n_samples":135,"n_pointer_only_licence":18,"note":"each paper (arXiv id) counted once, however many rows it is behind; this is the page-level figure"},"samples_row_weighted":{"n_ran":71,"n_unverified":64,"n_samples":135,"n_pointer_only_licence":18,"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"}}}