{"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/constrained-size-tensorflow-models-for","title":"Constrained-size Tensorflow Models for YouTube-8M Video Understanding Challenge","arxiv_id":"1808.06739","date":"2018-08-21","proceeding":null,"authors":["Tianqi Liu","Bo Liu"],"abstract":"This paper presents our 7th place solution to the second YouTube-8M video\nunderstanding competition which challenges participates to build a\nconstrained-size model to classify millions of YouTube videos into thousands of\nclasses. Our final model consists of four single models aggregated into one\ntensorflow graph. For each single model, we use the same network architecture\nas in the winning solution of the first YouTube-8M video understanding\ncompetition, namely Gated NetVLAD. We train the single models separately in\ntensorflow's default float32 precision, then replace weights with float16\nprecision and ensemble them in the evaluation and inference stages., achieving\n48.5% compression rate without loss of precision. Our best model achieved\n88.324% GAP on private leaderboard. The code is publicly available at\nhttps://github.com/boliu61/youtube-8m","url_abs":"http://arxiv.org/abs/1808.06739v3","url_pdf":"http://arxiv.org/pdf/1808.06739v3.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":"constrained-size-tensorflow-models-for","repo_url":"https://github.com/boliu61/youtube-8m","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"tf","reach":null},{"paper_slug":"constrained-size-tensorflow-models-for","repo_url":"https://github.com/TerenceLiu4444/Kaggle-solutions","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"tf","reach":null}],"tasks":[{"task_slug":"video-understanding","task_name":"Video Understanding"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":null,"mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}