{"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/deep-convolutional-acoustic-word-embeddings","title":"Deep convolutional acoustic word embeddings using word-pair side information","arxiv_id":"1510.01032","date":"2015-10-05","proceeding":null,"authors":["Herman Kamper","Weiran Wang","Karen Livescu"],"abstract":"Recent studies have been revisiting whole words as the basic modelling unit\nin speech recognition and query applications, instead of phonetic units. Such\nwhole-word segmental systems rely on a function that maps a variable-length\nspeech segment to a vector in a fixed-dimensional space; the resulting acoustic\nword embeddings need to allow for accurate discrimination between different\nword types, directly in the embedding space. We compare several old and new\napproaches in a word discrimination task. Our best approach uses side\ninformation in the form of known word pairs to train a Siamese convolutional\nneural network (CNN): a pair of tied networks that take two speech segments as\ninput and produce their embeddings, trained with a hinge loss that separates\nsame-word pairs and different-word pairs by some margin. A word classifier CNN\nperforms similarly, but requires much stronger supervision. Both types of CNNs\nyield large improvements over the best previously published results on the word\ndiscrimination task.","url_abs":"http://arxiv.org/abs/1510.01032v2","url_pdf":"http://arxiv.org/pdf/1510.01032v2.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":"deep-convolutional-acoustic-word-embeddings","repo_url":"https://github.com/kamperh/recipe_swbd_wordembeds","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"none","reach":{"status":"ok"}}],"tasks":[{"task_slug":"speech-recognition","task_name":"Speech Recognition"},{"task_slug":"word-embeddings","task_name":"Word Embeddings"},{"task_slug":"speech-recognition-1","task_name":"speech-recognition"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1510.01032","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}