{"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/deepmatching-hierarchical-deformable-dense","title":"DeepMatching: Hierarchical Deformable Dense Matching","arxiv_id":"1506.07656","date":"2015-06-25","proceeding":null,"authors":["Jerome Revaud","Philippe Weinzaepfel","Zaid Harchaoui","Cordelia Schmid"],"abstract":"We introduce a novel matching algorithm, called DeepMatching, to compute\ndense correspondences between images. DeepMatching relies on a hierarchical,\nmulti-layer, correlational architecture designed for matching images and was\ninspired by deep convolutional approaches. The proposed matching algorithm can\nhandle non-rigid deformations and repetitive textures and efficiently\ndetermines dense correspondences in the presence of significant changes between\nimages. We evaluate the performance of DeepMatching, in comparison with\nstate-of-the-art matching algorithms, on the Mikolajczyk (Mikolajczyk et al\n2005), the MPI-Sintel (Butler et al 2012) and the Kitti (Geiger et al 2013)\ndatasets. DeepMatching outperforms the state-of-the-art algorithms and shows\nexcellent results in particular for repetitive textures.We also propose a\nmethod for estimating optical flow, called DeepFlow, by integrating\nDeepMatching in the large displacement optical flow (LDOF) approach of Brox and\nMalik (2011). Compared to existing matching algorithms, additional robustness\nto large displacements and complex motion is obtained thanks to our matching\napproach. DeepFlow obtains competitive performance on public benchmarks for\noptical flow estimation.","url_abs":"http://arxiv.org/abs/1506.07656v2","url_pdf":"http://arxiv.org/pdf/1506.07656v2.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":"deepmatching-hierarchical-deformable-dense","repo_url":"https://github.com/vwegn/dm","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"tf","reach":{"status":"ok","spdx":"MIT"}}],"tasks":[{"task_slug":"dense-pixel-correspondence-estimation","task_name":"Dense Pixel Correspondence Estimation"},{"task_slug":"optical-flow-estimation","task_name":"Optical Flow Estimation"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/dense-pixel-correspondence-estimation-on","task":"Dense Pixel Correspondence Estimation","dataset":"HPatches","model":"DeepMatching*","rank_in_archive_order":4,"of":8,"metrics":{"Viewpoint I AEPE":"5.84","Viewpoint II AEPE":"4.63","Viewpoint III AEPE":"12.43","Viewpoint IV AEPE":"12.17","Viewpoint V AEPE":"22.55"},"uses_additional_data":false}],"syntology":{"syntology_url":null,"atlas_url":null,"mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}