{"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/read-watch-and-move-reinforcement-learning","title":"Read, Watch, and Move: Reinforcement Learning for Temporally Grounding Natural Language Descriptions in Videos","arxiv_id":"1901.06829","date":"2019-01-21","proceeding":null,"authors":["Dongliang He","Xiang Zhao","Jizhou Huang","Fu Li","Xiao Liu","Shilei Wen"],"abstract":"The task of video grounding, which temporally localizes a natural language\ndescription in a video, plays an important role in understanding videos.\nExisting studies have adopted strategies of sliding window over the entire\nvideo or exhaustively ranking all possible clip-sentence pairs in a\npre-segmented video, which inevitably suffer from exhaustively enumerated\ncandidates. To alleviate this problem, we formulate this task as a problem of\nsequential decision making by learning an agent which regulates the temporal\ngrounding boundaries progressively based on its policy. Specifically, we\npropose a reinforcement learning based framework improved by multi-task\nlearning and it shows steady performance gains by considering additional\nsupervised boundary information during training. Our proposed framework\nachieves state-of-the-art performance on ActivityNet'18 DenseCaption dataset\nand Charades-STA dataset while observing only 10 or less clips per video.","url_abs":"http://arxiv.org/abs/1901.06829v1","url_pdf":"http://arxiv.org/pdf/1901.06829v1.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":"read-watch-and-move-reinforcement-learning","repo_url":"https://github.com/WuJie1010/Temporally-language-grounding","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok"}}],"tasks":[{"task_slug":"decision-making","task_name":"Decision Making"},{"task_slug":"multi-task-learning","task_name":"Multi-Task Learning"},{"task_slug":"reinforcement-learning","task_name":"Reinforcement Learning"},{"task_slug":"reinforcement-learning-1","task_name":"Reinforcement Learning (RL)"},{"task_slug":"sentence","task_name":"Sentence"},{"task_slug":"sequential-decision-making","task_name":"Sequential Decision Making"},{"task_slug":"video-grounding","task_name":"Video Grounding"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1901.06829","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}