{"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/end-to-end-fine-grained-action-segmentation","title":"End-to-End Fine-Grained Action Segmentation and Recognition Using Conditional Random Field Models and Discriminative Sparse Coding","arxiv_id":"1801.09571","date":"2018-01-29","proceeding":null,"authors":["Effrosyni Mavroudi","Divya Bhaskara","Shahin Sefati","Haider Ali","René Vidal"],"abstract":"Fine-grained action segmentation and recognition is an important yet\nchallenging task. Given a long, untrimmed sequence of kinematic data, the task\nis to classify the action at each time frame and segment the time series into\nthe correct sequence of actions. In this paper, we propose a novel framework\nthat combines a temporal Conditional Random Field (CRF) model with a powerful\nframe-level representation based on discriminative sparse coding. We introduce\nan end-to-end algorithm for jointly learning the weights of the CRF model,\nwhich include action classification and action transition costs, as well as an\novercomplete dictionary of mid-level action primitives. This results in a CRF\nmodel that is driven by sparse coding features obtained using a discriminative\ndictionary that is shared among different actions and adapted to the task of\nstructured output learning. We evaluate our method on three surgical tasks\nusing kinematic data from the JIGSAWS dataset, as well as on a food preparation\ntask using accelerometer data from the 50 Salads dataset. Our results show that\nthe proposed method performs on par or better than state-of-the-art methods.","url_abs":"http://arxiv.org/abs/1801.09571v1","url_pdf":"http://arxiv.org/pdf/1801.09571v1.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":"action-classification","task_name":"Action Classification"},{"task_slug":"action-segmentation","task_name":"Action Segmentation"},{"task_slug":"time-series-1","task_name":"Time Series"},{"task_slug":"time-series","task_name":"Time Series Analysis"}],"methods":[{"method_slug":"crf","method_name":"CRF"}],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/action-segmentation-on-jigsaws","task":"Action Segmentation","dataset":"JIGSAWS","model":"SDL+SC-CRF","rank_in_archive_order":5,"of":7,"metrics":{"Edit Distance":"86.21"},"uses_additional_data":false}],"syntology":{"syntology_url":null,"atlas_url":"https://app.syntology.ai/?focus=1801.09571","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}