{"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/cross-modal-embeddings-for-video-and-audio","title":"Cross-modal Embeddings for Video and Audio Retrieval","arxiv_id":"1801.02200","date":"2018-01-07","proceeding":null,"authors":["Didac Surís","Amanda Duarte","Amaia Salvador","Jordi Torres","Xavier Giró-i-Nieto"],"abstract":"The increasing amount of online videos brings several opportunities for\ntraining self-supervised neural networks. The creation of large scale datasets\nof videos such as the YouTube-8M allows us to deal with this large amount of\ndata in manageable way. In this work, we find new ways of exploiting this\ndataset by taking advantage of the multi-modal information it provides. By\nmeans of a neural network, we are able to create links between audio and visual\ndocuments, by projecting them into a common region of the feature space,\nobtaining joint audio-visual embeddings. These links are used to retrieve audio\nsamples that fit well to a given silent video, and also to retrieve images that\nmatch a given a query audio. The results in terms of Recall@K obtained over a\nsubset of YouTube-8M videos show the potential of this unsupervised approach\nfor cross-modal feature learning. We train embeddings for both scales and\nassess their quality in a retrieval problem, formulated as using the feature\nextracted from one modality to retrieve the most similar videos based on the\nfeatures computed in the other modality.","url_abs":"http://arxiv.org/abs/1801.02200v1","url_pdf":"http://arxiv.org/pdf/1801.02200v1.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":"cross-modal-embeddings-for-video-and-audio","repo_url":"https://github.com/surisdi/youtube-8m","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":0,"framework":"tf","reach":{"status":"unanswered"}}],"tasks":[{"task_slug":"retrieval","task_name":"Retrieval"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1801.02200","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}