{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,5,4]],"date-time":"2026-05-04T10:10:02Z","timestamp":1777889402698,"version":"3.51.4"},"reference-count":59,"publisher":"IEEE","license":[{"start":{"date-parts":[[2025,10,19]],"date-time":"2025-10-19T00:00:00Z","timestamp":1760832000000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-029"},{"start":{"date-parts":[[2025,10,19]],"date-time":"2025-10-19T00:00:00Z","timestamp":1760832000000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-037"}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":[],"published-print":{"date-parts":[[2025,10,19]]},"DOI":"10.1109\/iccv51701.2025.02190","type":"proceedings-article","created":{"date-parts":[[2026,4,29]],"date-time":"2026-04-29T19:45:49Z","timestamp":1777491949000},"page":"23593-23603","source":"Crossref","is-referenced-by-count":0,"title":["TOGA: Temporally Grounded Open-Ended Video QA with Weak Supervision"],"prefix":"10.1109","author":[{"given":"Ayush","family":"Gupta","sequence":"first","affiliation":[{"name":"SRI"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Anirban","family":"Roy","sequence":"additional","affiliation":[{"name":"SRI"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Rama","family":"Chellappa","sequence":"additional","affiliation":[{"name":"Johns Hopkins University"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Nathaniel D.","family":"Bastian","sequence":"additional","affiliation":[{"name":"United States Military Academy"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Alvaro","family":"Velasquez","sequence":"additional","affiliation":[{"name":"University of Colorado Boulder"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Susmit","family":"Jha","sequence":"additional","affiliation":[{"name":"SRI"}],"role":[{"role":"author","vocabulary":"crossref"}]}],"member":"263","reference":[{"key":"ref1","doi-asserted-by":"publisher","DOI":"10.1109\/ICCV.2015.279"},{"key":"ref2","first-page":"190","article-title":"Collecting highly parallel data for paraphrase evaluation","volume-title":"Proceedings of the 49th Annual Meeting of the Association for Computational Linguistics: Human Language Technologies","author":"Chen"},{"key":"ref3","doi-asserted-by":"publisher","DOI":"10.52202\/079017-0900"},{"key":"ref4","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52733.2024.01265"},{"key":"ref5","author":"Cheng","year":"2024","journal-title":"Videollama 2: Advancing spatial-temporal modeling and audio understanding in video1 lms"},{"key":"ref6","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2015.7298698"},{"key":"ref7","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2020.emnlp-main.61"},{"key":"ref8","doi-asserted-by":"publisher","DOI":"10.1109\/ICCV.2019.00630"},{"key":"ref9","doi-asserted-by":"publisher","DOI":"10.1609\/aaai.v34i07.6713"},{"key":"ref10","doi-asserted-by":"publisher","DOI":"10.1007\/s11263-018-1116-0"},{"key":"ref11","doi-asserted-by":"publisher","DOI":"10.21236\/ADA531042"},{"key":"ref12","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-031-73039-9_12"},{"key":"ref13","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2017.149"},{"key":"ref14","doi-asserted-by":"publisher","DOI":"10.1007\/s11263-019-01189-x"},{"key":"ref15","doi-asserted-by":"publisher","DOI":"10.1007\/s11263-019-01189-x"},{"key":"ref16","author":"Jiang","year":"2023","journal-title":"Mistral 7b"},{"key":"ref17","doi-asserted-by":"publisher","DOI":"10.1109\/cvpr52733.2024.01300"},{"key":"ref18","doi-asserted-by":"publisher","DOI":"10.1109\/ICASSP39728.2021.9413376"},{"key":"ref19","doi-asserted-by":"publisher","DOI":"10.1109\/ICCV51070.2023.00288"},{"key":"ref20","doi-asserted-by":"publisher","DOI":"10.1109\/ICCV.2017.83"},{"key":"ref21","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/d18-1167"},{"key":"ref22","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2020.acl-main.730"},{"key":"ref23","doi-asserted-by":"publisher","DOI":"10.1007\/s11432-024-4321-9"},{"key":"ref24","doi-asserted-by":"publisher","DOI":"10.1145\/3503161.3548035"},{"key":"ref25","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2024.emnlp-main.342"},{"key":"ref26","doi-asserted-by":"publisher","DOI":"10.52202\/075280-1516"},{"key":"ref27","first-page":"5","author":"Loshchilov","year":"2017","journal-title":"Decoupled weight decay regularization"},{"key":"ref28","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2024.acl-long.679"},{"key":"ref29","first-page":"8","author":"Meinardus","year":"2024","journal-title":"The surprising effectiveness of multimodal large language models for video moment retrieval"},{"key":"ref30","first-page":"4","author":"Micheletti","year":"2024","journal-title":"Exploration of masked and causal language modelling for text generation"},{"key":"ref31","doi-asserted-by":"publisher","DOI":"10.1145\/219717.219748"},{"key":"ref32","article-title":"OpenAI","year":"2024","journal-title":"Gpt-4 technical report"},{"key":"ref33","doi-asserted-by":"publisher","DOI":"10.52202\/079017-3792"},{"key":"ref34","author":"Radford","year":"2021","journal-title":"Learning transferable visual models from natural language supervision"},{"key":"ref35","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR42600.2020.01044"},{"key":"ref36","doi-asserted-by":"publisher","DOI":"10.1609\/aaai.v31i1.11164"},{"key":"ref37","doi-asserted-by":"publisher","DOI":"10.1109\/ICCV51070.2023.01092"},{"key":"ref38","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2016.501"},{"key":"ref39","first-page":"2","article-title":"Gemini Team","year":"2024","journal-title":"Gemini: A family of highly capable multimodal models"},{"key":"ref40","author":"Touvron","year":"2023","journal-title":"Llama: Open and efficient foundation language models"},{"key":"ref41","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2025.findings-emnlp.50"},{"key":"ref42","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-031-72989-8_4"},{"key":"ref43","first-page":"2","author":"Wang","year":"2024","journal-title":"Videotree: Adaptive tree-based video representation for 11 m reasoning on long videos"},{"key":"ref44","first-page":"5","author":"Wu","year":"2024","journal-title":"Star: A benchmark for situated reasoning in real-world videos"},{"key":"ref45","doi-asserted-by":"publisher","DOI":"10.1109\/cvpr52734.2025.01284"},{"key":"ref46","author":"Xiao","year":"2020","journal-title":"Audiovisual slowfast networks for video recognition"},{"key":"ref47","doi-asserted-by":"publisher","DOI":"10.1109\/cvpr46437.2021.00965"},{"key":"ref48","doi-asserted-by":"publisher","DOI":"10.1007\/s11263-025-02385-8"},{"key":"ref49","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52733.2024.01254"},{"key":"ref50","doi-asserted-by":"publisher","DOI":"10.1145\/3123266.3123427"},{"key":"ref51","doi-asserted-by":"publisher","DOI":"10.1145\/3123266.3123427"},{"key":"ref52","article-title":"Zero-shot video question answering via frozen bidirectional language models","author":"Yang","year":"2022","journal-title":"NeurIPS"},{"key":"ref53","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52688.2022.01595"},{"key":"ref54","article-title":"Self-chained image-language model for video localization and question answering","author":"Yu","year":"2023","journal-title":"NeurIPS"},{"key":"ref55","doi-asserted-by":"publisher","DOI":"10.1609\/aaai.v33i01.33019127"},{"key":"ref56","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2024.emnlp-main.1209"},{"key":"ref57","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2023.emnlp-demo.49"},{"key":"ref58","first-page":"6","author":"Zhang","year":"2023","journal-title":"Llama-adapter: Efficient fine-tuning of language models with zero-init attention"},{"key":"ref59","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2022.emnlp-main.432"}],"event":{"name":"2025 IEEE\/CVF International Conference on Computer Vision (ICCV)","location":"Honolulu, HI, USA","start":{"date-parts":[[2025,10,19]]},"end":{"date-parts":[[2025,10,25]]}},"container-title":["2025 IEEE\/CVF International Conference on Computer Vision (ICCV)"],"original-title":[],"link":[{"URL":"http:\/\/xplorestaging.ieee.org\/ielx8\/11443115\/11443287\/11443295.pdf?arnumber=11443295","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,5,1]],"date-time":"2026-05-01T05:12:19Z","timestamp":1777612339000},"score":1,"resource":{"primary":{"URL":"https:\/\/ieeexplore.ieee.org\/document\/11443295\/"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2025,10,19]]},"references-count":59,"URL":"https:\/\/doi.org\/10.1109\/iccv51701.2025.02190","relation":{},"subject":[],"published":{"date-parts":[[2025,10,19]]}}}