{"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/a-large-dataset-for-improving-patch-matching","title":"A Large Dataset for Improving Patch Matching","arxiv_id":"1801.01466","date":"2018-01-04","proceeding":null,"authors":["Rahul Mitra","Nehal Doiphode","Utkarsh Gautam","Sanath Narayan","Shuaib Ahmed","Sharat Chandran","Arjun Jain"],"abstract":"We propose a new dataset for learning local image descriptors which can be\nused for significantly improved patch matching. Our proposed dataset consists\nof an order of magnitude more number of scenes, images, and positive and\nnegative correspondences compared to the currently available Multi-View Stereo\n(MVS) dataset from Brown et al. The new dataset also has better coverage of the\noverall viewpoint, scale, and lighting changes in comparison to the MVS\ndataset. Our dataset also provides supplementary information like RGB patches\nwith scale and rotations values, and intrinsic and extrinsic camera parameters\nwhich as shown later can be used to customize training data as per application.\nWe train an existing state-of-the-art model on our dataset and evaluate on\npublicly available benchmarks such as HPatches dataset and Strecha et\nal.\\cite{strecha} to quantify the image descriptor performance. Experimental\nevaluations show that the descriptors trained using our proposed dataset\noutperform the current state-of-the-art descriptors trained on MVS by 8%, 4%\nand 10% on matching, verification and retrieval tasks respectively on the\nHPatches dataset. Similarly on the Strecha dataset, we see an improvement of\n3-5% for the matching task in non-planar scenes.","url_abs":"http://arxiv.org/abs/1801.01466v3","url_pdf":"http://arxiv.org/pdf/1801.01466v3.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":"a-large-dataset-for-improving-patch-matching","repo_url":"https://github.com/rmitra/PS-Dataset","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"none","reach":null}],"tasks":[{"task_slug":"patch-matching","task_name":"Patch Matching"},{"task_slug":"retrieval","task_name":"Retrieval"}],"methods":[],"datasets_introduced":[{"slug":"photosynth","name":"PhotoSynth","full_name":null}],"methods_introduced":[],"results":[],"syntology":{"atlas_url":null,"mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}