{"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/190408103","title":"Multi-Scale Geometric Consistency Guided Multi-View Stereo","arxiv_id":"1904.08103","date":"2019-04-17","proceeding":"CVPR 2019 6","authors":["Qingshan Xu","Wenbing Tao"],"abstract":"In this paper, we propose an efficient multi-scale geometric consistency\nguided multi-view stereo method for accurate and complete depth map estimation.\nWe first present our basic multi-view stereo method with Adaptive Checkerboard\nsampling and Multi-Hypothesis joint view selection (ACMH). It leverages\nstructured region information to sample better candidate hypotheses for\npropagation and infer the aggregation view subset at each pixel. For the depth\nestimation of low-textured areas, we further propose to combine ACMH with\nmulti-scale geometric consistency guidance (ACMM) to obtain the reliable depth\nestimates for low-textured areas at coarser scales and guarantee that they can\nbe propagated to finer scales. To correct the erroneous estimates propagated\nfrom the coarser scales, we present a novel detail restorer. Experiments on\nextensive datasets show our method achieves state-of-the-art performance,\nrecovering the depth estimation not only in low-textured areas but also in\ndetails.","url_abs":"http://arxiv.org/abs/1904.08103v1","url_pdf":"http://arxiv.org/pdf/1904.08103v1.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":"depth-estimation","task_name":"Depth Estimation"},{"task_slug":"multi-view-3d-reconstruction","task_name":"Multi-View 3D Reconstruction"},{"task_slug":"point-clouds","task_name":"Point Clouds"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/multi-view-3d-reconstruction-on-eth3d","task":"Multi-View 3D Reconstruction","dataset":"ETH3D","model":"ACMM","rank_in_archive_order":5,"of":5,"metrics":{"F1 score":"80.78"},"uses_additional_data":false},{"leaderboard":"/sota/point-clouds-on-tanks-and-temples","task":"Point Clouds","dataset":"Tanks and Temples","model":"ACMM","rank_in_archive_order":10,"of":21,"metrics":{"Mean F1 (Advanced)":"34.02","Mean F1 (Intermediate)":"57.27"},"uses_additional_data":false}],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1904.08103","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}