{"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/adadepth-unsupervised-content-congruent","title":"AdaDepth: Unsupervised Content Congruent Adaptation for Depth Estimation","arxiv_id":"1803.01599","date":"2018-03-05","proceeding":"CVPR 2018 6","authors":["Jogendra Nath Kundu","Phani Krishna Uppala","Anuj Pahuja","R. Venkatesh Babu"],"abstract":"Supervised deep learning methods have shown promising results for the task of\nmonocular depth estimation; but acquiring ground truth is costly, and prone to\nnoise as well as inaccuracies. While synthetic datasets have been used to\ncircumvent above problems, the resultant models do not generalize well to\nnatural scenes due to the inherent domain shift. Recent adversarial approaches\nfor domain adaption have performed well in mitigating the differences between\nthe source and target domains. But these methods are mostly limited to a\nclassification setup and do not scale well for fully-convolutional\narchitectures. In this work, we propose AdaDepth - an unsupervised domain\nadaptation strategy for the pixel-wise regression task of monocular depth\nestimation. The proposed approach is devoid of above limitations through a)\nadversarial learning and b) explicit imposition of content consistency on the\nadapted target representation. Our unsupervised approach performs competitively\nwith other established approaches on depth estimation tasks and achieves\nstate-of-the-art results in a semi-supervised setting.","url_abs":"http://arxiv.org/abs/1803.01599v2","url_pdf":"http://arxiv.org/pdf/1803.01599v2.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":"domain-adaptation","task_name":"Domain Adaptation"},{"task_slug":"monocular-depth-estimation","task_name":"Monocular Depth Estimation"},{"task_slug":"unsupervised-domain-adaptation","task_name":"Unsupervised Domain Adaptation"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/unsupervised-domain-adaptation-on-virtual-2","task":"Unsupervised Domain Adaptation","dataset":"virtual KITTI to KITTI (MDE)","model":"AdaDepth","rank_in_archive_order":4,"of":4,"metrics":{"RMSE ":"6.251"},"uses_additional_data":false}],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1803.01599","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}