{"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/deep-reinforcement-learning-for-surgical","title":"Deep Reinforcement Learning for Surgical Gesture Segmentation and Classification","arxiv_id":"1806.08089","date":"2018-06-21","proceeding":null,"authors":["Daochang Liu","Tingting Jiang"],"abstract":"Recognition of surgical gesture is crucial for surgical skill assessment and\nefficient surgery training. Prior works on this task are based on either\nvariant graphical models such as HMMs and CRFs, or deep learning models such as\nRecurrent Neural Networks and Temporal Convolutional Networks. Most of the\ncurrent approaches usually suffer from over-segmentation and therefore low\nsegment-level edit scores. In contrast, we present an essentially different\nmethodology by modeling the task as a sequential decision-making process. An\nintelligent agent is trained using reinforcement learning with hierarchical\nfeatures from a deep model. Temporal consistency is integrated into our action\ndesign and reward mechanism to reduce over-segmentation errors. Experiments on\nJIGSAWS dataset demonstrate that the proposed method performs better than\nstate-of-the-art methods in terms of the edit score and on par in frame-wise\naccuracy. Our code will be released later.","url_abs":"http://arxiv.org/abs/1806.08089v1","url_pdf":"http://arxiv.org/pdf/1806.08089v1.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":"deep-reinforcement-learning-for-surgical","repo_url":"https://github.com/Finspire13/RL-Surgical-Gesture-Segmentation","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":null}],"tasks":[{"task_slug":"action-segmentation","task_name":"Action Segmentation"},{"task_slug":"classification-1","task_name":"Classification"},{"task_slug":"decision-making","task_name":"Decision Making"},{"task_slug":"deep-reinforcement-learning","task_name":"Deep Reinforcement Learning"},{"task_slug":"classification","task_name":"General Classification"},{"task_slug":"reinforcement-learning","task_name":"Reinforcement Learning"},{"task_slug":"reinforcement-learning-1","task_name":"Reinforcement Learning (RL)"},{"task_slug":"segmentation","task_name":"Segmentation"},{"task_slug":"sequential-decision-making","task_name":"Sequential Decision Making"},{"task_slug":"surgical-gesture-recognition","task_name":"Surgical Gesture Recognition"},{"task_slug":"reinforcement-learning-2","task_name":"reinforcement-learning"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/action-segmentation-on-jigsaws","task":"Action Segmentation","dataset":"JIGSAWS","model":"RL (full)","rank_in_archive_order":3,"of":7,"metrics":{"Accuracy":"81.43","Edit Distance":"87.96","F1@10":"92.0","F1@25":"90.5","F1@50":"82.2"},"uses_additional_data":false}],"syntology":{"syntology_url":null,"atlas_url":null,"mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}