{"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/learning-exploration-policies-for-navigation","title":"Learning Exploration Policies for Navigation","arxiv_id":"1903.01959","date":"2019-03-05","proceeding":"ICLR 2019 5","authors":["Tao Chen","Saurabh Gupta","Abhinav Gupta"],"abstract":"Numerous past works have tackled the problem of task-driven navigation. But,\nhow to effectively explore a new environment to enable a variety of down-stream\ntasks has received much less attention. In this work, we study how agents can\nautonomously explore realistic and complex 3D environments without the context\nof task-rewards. We propose a learning-based approach and investigate different\npolicy architectures, reward functions, and training paradigms. We find that\nthe use of policies with spatial memory that are bootstrapped with imitation\nlearning and finally finetuned with coverage rewards derived purely from\non-board sensors can be effective at exploring novel environments. We show that\nour learned exploration policies can explore better than classical approaches\nbased on geometry alone and generic learning-based exploration techniques.\nFinally, we also show how such task-agnostic exploration can be used for\ndown-stream tasks. Code and Videos are available at:\nhttps://sites.google.com/view/exploration-for-nav.","url_abs":"http://arxiv.org/abs/1903.01959v1","url_pdf":"http://arxiv.org/pdf/1903.01959v1.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":"learning-exploration-policies-for-navigation","repo_url":"https://github.com/taochenshh/exp4nav","is_official":1,"mentioned_in_paper":0,"mentioned_in_github":0,"framework":"pytorch","reach":{"status":"ok"}},{"paper_slug":"learning-exploration-policies-for-navigation","repo_url":"https://github.com/s-gupta/map-plan-baseline","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"none","reach":null}],"tasks":[{"task_slug":"efficient-exploration","task_name":"Efficient Exploration"},{"task_slug":"general-reinforcement-learning","task_name":"General Reinforcement Learning"},{"task_slug":"imitation-learning","task_name":"Imitation Learning"},{"task_slug":"visual-navigation","task_name":"Visual Navigation"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1903.01959","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}