{"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/real-time-seamless-single-shot-6d-object-pose","title":"Real-Time Seamless Single Shot 6D Object Pose Prediction","arxiv_id":"1711.08848","date":"2017-11-24","proceeding":"CVPR 2018 6","authors":["Bugra Tekin","Sudipta N. Sinha","Pascal Fua"],"abstract":"We propose a single-shot approach for simultaneously detecting an object in\nan RGB image and predicting its 6D pose without requiring multiple stages or\nhaving to examine multiple hypotheses. Unlike a recently proposed single-shot\ntechnique for this task (Kehl et al., ICCV'17) that only predicts an\napproximate 6D pose that must then be refined, ours is accurate enough not to\nrequire additional post-processing. As a result, it is much faster - 50 fps on\na Titan X (Pascal) GPU - and more suitable for real-time processing. The key\ncomponent of our method is a new CNN architecture inspired by the YOLO network\ndesign that directly predicts the 2D image locations of the projected vertices\nof the object's 3D bounding box. The object's 6D pose is then estimated using a\nPnP algorithm.\n  For single object and multiple object pose estimation on the LINEMOD and\nOCCLUSION datasets, our approach substantially outperforms other recent\nCNN-based approaches when they are all used without post-processing. During\npost-processing, a pose refinement step can be used to boost the accuracy of\nthe existing methods, but at 10 fps or less, they are much slower than our\nmethod.","url_abs":"http://arxiv.org/abs/1711.08848v5","url_pdf":"http://arxiv.org/pdf/1711.08848v5.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":"real-time-seamless-single-shot-6d-object-pose","repo_url":"https://github.com/Microsoft/singleshotpose","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"unanswered"}},{"paper_slug":"real-time-seamless-single-shot-6d-object-pose","repo_url":"https://github.com/Yongjjun/singleshotpose","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"unanswered"}},{"paper_slug":"real-time-seamless-single-shot-6d-object-pose","repo_url":"https://github.com/a2824256/singleshotpose_imp","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"unanswered"}},{"paper_slug":"real-time-seamless-single-shot-6d-object-pose","repo_url":"https://github.com/hz-ants/obtain-an-object-mesh-and-create-labels","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"unanswered"}},{"paper_slug":"real-time-seamless-single-shot-6d-object-pose","repo_url":"https://github.com/hz-ants/yolo-6d","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"unanswered"}}],"tasks":[{"task_slug":"6d-pose-estimation","task_name":"6D Pose Estimation using RGB"},{"task_slug":"drone-pose-estimation","task_name":"Drone Pose Estimation"},{"task_slug":null,"task_name":"GPU"},{"task_slug":"object","task_name":"Object"},{"task_slug":"pose-estimation","task_name":"Pose Estimation"},{"task_slug":"pose-prediction","task_name":"Pose Prediction"},{"task_slug":"prediction","task_name":"Prediction"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/6d-pose-estimation-on-linemod","task":"6D Pose Estimation using RGB","dataset":"LineMOD","model":"Single-shot Deep CNN","rank_in_archive_order":18,"of":22,"metrics":{"Accuracy":"90.37%","Mean ADD":"55.95","Mean IoU":"99.92"},"uses_additional_data":false},{"leaderboard":"/sota/6d-pose-estimation-on-occlusion","task":"6D Pose Estimation using RGB","dataset":"OCCLUSION","model":"Single-shot deep CNN","rank_in_archive_order":1,"of":2,"metrics":{"MAP":"0.48"},"uses_additional_data":false}],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1711.08848","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}