{"id":{"repo_id":"mit","oai_identifier":"oai:dspace.mit.edu:1721.1/144497"},"canonical_url":"https://search.dev.ndltd.org/etd/mit/oai:dspace.mit.edu:1721.1/144497","repository":{"repo_id":"mit","name":"MIT","base_url":"https://dspace.mit.edu/oai/request"},"display":{"title":"Inferring Structured World Models from Videos","abstract":"Advances in reinforcement learning have allowed agents to learn a variety of board games and video games at superhuman levels. Unlike humans - which can generalize to a wide range of tasks with very little experience - these algorithms typically need vast number of experience replays to perform at the same level. In this thesis, we propose a model-based reinforcement learning approach that represents the environment using an explicit symbolic model in the form of a domain-specific language (DSL) that represents the world as a set of discrete objects with underlying latent properties that govern their dynamical interactions. We present a novel, neurally guided, on-line inference technique to recover the structured world representation from raw video observations, with the intent to be used for downstream model-based planning. We qualitatively evaluate our inference performance on classical Atari games, as well as on physics-based mobile games.","abstract_html":"Advances in reinforcement learning have allowed agents to learn a variety of board games and video games at superhuman levels. Unlike humans - which can generalize to a wide range of tasks with very little experience - these algorithms typically need vast number of experience replays to perform at the same level. In this thesis, we propose a model-based reinforcement learning approach that represents the environment using an explicit symbolic model in the form of a domain-specific language (DSL) that represents the world as a set of discrete objects with underlying latent properties that govern their dynamical interactions. We present a novel, neurally guided, on-line inference technique to recover the structured world representation from raw video observations, with the intent to be used for downstream model-based planning. We qualitatively evaluate our inference performance on classical Atari games, as well as on physics-based mobile games.","abstract_has_math":false,"creators":["Kapur, Shreyas"],"institution":"Massachusetts Institute of Technology","degree_name":"Master","degree_level":null,"degree_discipline":null,"degree_department":"Massachusetts Institute of Technology. Department of Electrical Engineering and Computer Science","school":null,"contributors":[],"advisors":["Tenenbaum, Joshua B."],"committee_chairs":[],"committee_members":[],"year":2022,"date_issued":"2022-05","date_published":"2022-05","updated_at":"2026-07-22T22:21:13Z","subjects":[],"languages":[],"rights":["In Copyright - Educational Use Permitted","Copyright MIT"],"rights_urls":["http://rightsstatements.org/page/InC-EDU/1.0/"],"identifier_entries":[]},"links":{"outbound_url":"https://hdl.handle.net/1721.1/144497","outbound_label":"Handle","outbound_source":"dc:identifier.uri"},"metadata_groups":[{"id":"people","label":"People","entries":[{"key":"dc:contributor.advisor","label":"Advisor","values":["Tenenbaum, Joshua B."]},{"key":"dc:contributor.department","label":"Department","values":["Massachusetts Institute of Technology. Department of Electrical Engineering and Computer Science"]},{"key":"dc:creator","label":"Author","values":["Kapur, Shreyas"]}]},{"id":"academic_context","label":"Academic Context","entries":[{"key":"dc:date.accessioned","label":"Dc Date Accessioned","values":["2022-08-29T15:51:35Z"]},{"key":"dc:date.available","label":"Dc Date Available","values":["2022-08-29T15:51:35Z"]},{"key":"dc:date.issued","label":"Date","values":["2022-05"]},{"key":"dc:publisher","label":"Institution","values":["Massachusetts Institute of Technology"]},{"key":"dc:type","label":"Dc Type","values":["Thesis"]},{"key":"thesis:degree_name","label":"Degree Name","values":["Master","Master of Engineering in Electrical Engineering and Computer Science"]}]},{"id":"language_rights","label":"Language and Rights","entries":[{"key":"dc:rights","label":"Dc Rights","values":["In Copyright - Educational Use Permitted","Copyright MIT"]},{"key":"dc:rights.uri","label":"Rights URI","values":["http://rightsstatements.org/page/InC-EDU/1.0/"]}]},{"id":"identifiers","label":"Identifiers","entries":[{"key":"dc:identifier.uri","label":"Identifier URI","values":["https://hdl.handle.net/1721.1/144497"]}]},{"id":"additional","label":"Additional Metadata","entries":[{"key":"dc:description.abstract","label":"Abstract","values":["Advances in reinforcement learning have allowed agents to learn a variety of board games and video games at superhuman levels. Unlike humans - which can generalize to a wide range of tasks with very little experience - these algorithms typically need vast number of experience replays to perform at the same level. In this thesis, we propose a model-based reinforcement learning approach that represents the environment using an explicit symbolic model in the form of a domain-specific language (DSL) that represents the world as a set of discrete objects with underlying latent properties that govern their dynamical interactions. We present a novel, neurally guided, on-line inference technique to recover the structured world representation from raw video observations, with the intent to be used for downstream model-based planning. We qualitatively evaluate our inference performance on classical Atari games, as well as on physics-based mobile games."]},{"key":"dc:description.degree","label":"Dc Description Degree","values":["M.Eng."]},{"key":"dc:title","label":"Title","values":["Inferring Structured World Models from Videos"]}]}],"canonical_facts":{"dc:contributor.advisor":["Tenenbaum, Joshua B."],"dc:contributor.department":["Massachusetts Institute of Technology. Department of Electrical Engineering and Computer Science"],"dc:creator":["Kapur, Shreyas"],"dc:date.accessioned":["2022-08-29T15:51:35Z"],"dc:date.available":["2022-08-29T15:51:35Z"],"dc:date.issued":["2022-05"],"dc:description.abstract":["Advances in reinforcement learning have allowed agents to learn a variety of board games and video games at superhuman levels. Unlike humans - which can generalize to a wide range of tasks with very little experience - these algorithms typically need vast number of experience replays to perform at the same level. In this thesis, we propose a model-based reinforcement learning approach that represents the environment using an explicit symbolic model in the form of a domain-specific language (DSL) that represents the world as a set of discrete objects with underlying latent properties that govern their dynamical interactions. We present a novel, neurally guided, on-line inference technique to recover the structured world representation from raw video observations, with the intent to be used for downstream model-based planning. We qualitatively evaluate our inference performance on classical Atari games, as well as on physics-based mobile games."],"dc:description.degree":["M.Eng."],"dc:identifier.uri":["https://hdl.handle.net/1721.1/144497"],"dc:publisher":["Massachusetts Institute of Technology"],"dc:rights":["In Copyright - Educational Use Permitted","Copyright MIT"],"dc:rights.uri":["http://rightsstatements.org/page/InC-EDU/1.0/"],"dc:title":["Inferring Structured World Models from Videos"],"dc:type":["Thesis"],"thesis:degree_name":["Master","Master of Engineering in Electrical Engineering and Computer Science"]},"updated_at":"2026-07-22T22:21:13Z"}