{"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/geometry-aware-symmetric-domain-adaptation","title":"Geometry-Aware Symmetric Domain Adaptation for Monocular Depth Estimation","arxiv_id":"1904.01870","date":"2019-04-03","proceeding":"CVPR 2019 6","authors":["Shanshan Zhao","Huan Fu","Mingming Gong","DaCheng Tao"],"abstract":"Supervised depth estimation has achieved high accuracy due to the advanced\ndeep network architectures. Since the groundtruth depth labels are hard to\nobtain, recent methods try to learn depth estimation networks in an\nunsupervised way by exploring unsupervised cues, which are effective but less\nreliable than true labels. An emerging way to resolve this dilemma is to\ntransfer knowledge from synthetic images with ground truth depth via domain\nadaptation techniques. However, these approaches overlook specific geometric\nstructure of the natural images in the target domain (i.e., real data), which\nis important for high-performing depth prediction. Motivated by the\nobservation, we propose a geometry-aware symmetric domain adaptation framework\n(GASDA) to explore the labels in the synthetic data and epipolar geometry in\nthe real data jointly. Moreover, by training two image style translators and\ndepth estimators symmetrically in an end-to-end network, our model achieves\nbetter image style transfer and generates high-quality depth maps. The\nexperimental results demonstrate the effectiveness of our proposed method and\ncomparable performance against the state-of-the-art. Code will be publicly\navailable at: https://github.com/sshan-zhao/GASDA.","url_abs":"http://arxiv.org/abs/1904.01870v1","url_pdf":"http://arxiv.org/pdf/1904.01870v1.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":"geometry-aware-symmetric-domain-adaptation","repo_url":"https://github.com/sshan-zhao/GASDA","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":0,"framework":"pytorch","reach":{"status":"ok"}}],"tasks":[{"task_slug":"depth-estimation","task_name":"Depth Estimation"},{"task_slug":"depth-prediction","task_name":"Depth Prediction"},{"task_slug":"domain-adaptation","task_name":"Domain Adaptation"},{"task_slug":"monocular-depth-estimation","task_name":"Monocular Depth Estimation"},{"task_slug":"style-transfer","task_name":"Style Transfer"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/monocular-depth-estimation-on-kitti-eigen","task":"Monocular Depth Estimation","dataset":"KITTI Eigen split","model":"GASDA","rank_in_archive_order":78,"of":79,"metrics":{"absolute relative error":"0.149"},"uses_additional_data":false}],"syntology":{"syntology_url":"https://syntology.ai/paper/1904.01870","atlas_url":"https://app.syntology.ai/?focus=1904.01870","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}