{"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/playing-doom-with-slam-augmented-deep","title":"Playing Doom with SLAM-Augmented Deep Reinforcement Learning","arxiv_id":"1612.00380","date":"2016-12-01","proceeding":null,"authors":["Shehroze Bhatti","Alban Desmaison","Ondrej Miksik","Nantas Nardelli","N. Siddharth","Philip H. S. Torr"],"abstract":"A number of recent approaches to policy learning in 2D game domains have been\nsuccessful going directly from raw input images to actions. However when\nemployed in complex 3D environments, they typically suffer from challenges\nrelated to partial observability, combinatorial exploration spaces, path\nplanning, and a scarcity of rewarding scenarios. Inspired from prior work in\nhuman cognition that indicates how humans employ a variety of semantic concepts\nand abstractions (object categories, localisation, etc.) to reason about the\nworld, we build an agent-model that incorporates such abstractions into its\npolicy-learning framework. We augment the raw image input to a Deep Q-Learning\nNetwork (DQN), by adding details of objects and structural elements\nencountered, along with the agent's localisation. The different components are\nautomatically extracted and composed into a topological representation using\non-the-fly object detection and 3D-scene reconstruction.We evaluate the\nefficacy of our approach in Doom, a 3D first-person combat game that exhibits a\nnumber of challenges discussed, and show that our augmented framework\nconsistently learns better, more effective policies.","url_abs":"http://arxiv.org/abs/1612.00380v1","url_pdf":"http://arxiv.org/pdf/1612.00380v1.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":"playing-doom-with-slam-augmented-deep","repo_url":"https://github.com/shehroze37/Augmented-Deep-Reinforcement-Learning-for-3D-environments","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":0,"framework":"none","reach":{"status":"ok"}}],"tasks":[{"task_slug":"deep-reinforcement-learning","task_name":"Deep Reinforcement Learning"},{"task_slug":"object-detection","task_name":"Object Detection"},{"task_slug":"q-learning","task_name":"Q-Learning"},{"task_slug":"reinforcement-learning","task_name":"Reinforcement Learning"},{"task_slug":"reinforcement-learning-1","task_name":"Reinforcement Learning (RL)"},{"task_slug":"object-detection-1","task_name":"object-detection"},{"task_slug":"reinforcement-learning-2","task_name":"reinforcement-learning"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1612.00380","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}