{"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/3d-hand-pose-estimation-using-simulation-and","title":"3D Hand Pose Estimation using Simulation and Partial-Supervision with a Shared Latent Space","arxiv_id":"1807.05380","date":"2018-07-14","proceeding":"British Machine Vision Conference 2018 2018 7","authors":["Masoud Abdi","Ehsan Abbasnejad","Chee Peng Lim","Saeid Nahavandi"],"abstract":"Tremendous amounts of expensive annotated data are a vital ingredient for\nstate-of-the-art 3d hand pose estimation. Therefore, synthetic data has been\npopularized as annotations are automatically available. However, models trained\nonly with synthetic samples do not generalize to real data, mainly due to the\ngap between the distribution of synthetic and real data. In this paper, we\npropose a novel method that seeks to predict the 3d position of the hand using\nboth synthetic and partially-labeled real data. Accordingly, we form a shared\nlatent space between three modalities: synthetic depth image, real depth image,\nand pose. We demonstrate that by carefully learning the shared latent space, we\ncan find a regression model that is able to generalize to real data. As such,\nwe show that our method produces accurate predictions in both semi-supervised\nand unsupervised settings. Additionally, the proposed model is capable of\ngenerating novel, meaningful, and consistent samples from all of the three\ndomains. We evaluate our method qualitatively and quantitively on two highly\ncompetitive benchmarks (i.e., NYU and ICVL) and demonstrate its superiority\nover the state-of-the-art methods. The source code will be made available at\nhttps://github.com/masabdi/LSPS.","url_abs":"http://arxiv.org/abs/1807.05380v1","url_pdf":"http://arxiv.org/pdf/1807.05380v1.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":"3d-hand-pose-estimation-using-simulation-and","repo_url":"https://github.com/masabdi/LSPS","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"pytorch","reach":null}],"tasks":[{"task_slug":"3d-hand-pose-estimation","task_name":"3D Hand Pose Estimation"},{"task_slug":"hand-pose-estimation","task_name":"Hand Pose Estimation"},{"task_slug":"pose-estimation","task_name":"Pose Estimation"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"syntology_url":"https://syntology.ai/paper/1807.05380","atlas_url":"https://app.syntology.ai/?focus=1807.05380","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}