{"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/speech-model-pre-training-for-end-to-end","title":"Speech Model Pre-training for End-to-End Spoken Language Understanding","arxiv_id":"1904.03670","date":"2019-04-07","proceeding":null,"authors":["Loren Lugosch","Mirco Ravanelli","Patrick Ignoto","Vikrant Singh Tomar","Yoshua Bengio"],"abstract":"Whereas conventional spoken language understanding (SLU) systems map speech to text, and then text to intent, end-to-end SLU systems map speech directly to intent through a single trainable model. Achieving high accuracy with these end-to-end models without a large amount of training data is difficult. We propose a method to reduce the data requirements of end-to-end SLU in which the model is first pre-trained to predict words and phonemes, thus learning good features for SLU. We introduce a new SLU dataset, Fluent Speech Commands, and show that our method improves performance both when the full dataset is used for training and when only a small subset is used. We also describe preliminary experiments to gauge the model's ability to generalize to new phrases not heard during training.","url_abs":"https://arxiv.org/abs/1904.03670v2","url_pdf":"https://arxiv.org/pdf/1904.03670v2.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":"speech-model-pre-training-for-end-to-end","repo_url":"https://github.com/dscripka/openwakeword","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok","spdx":"Apache-2.0"}},{"paper_slug":"speech-model-pre-training-for-end-to-end","repo_url":"https://github.com/lorenlugosch/end-to-end-SLU","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok","spdx":"Apache-2.0"}}],"tasks":[{"task_slug":"speech-to-text","task_name":"Speech-to-Text"},{"task_slug":"spoken-language-understanding","task_name":"Spoken Language Understanding"}],"methods":[],"datasets_introduced":[{"slug":"fluent-speech-commands","name":"Fluent Speech Commands","full_name":""}],"methods_introduced":[],"results":[{"leaderboard":"/sota/spoken-language-understanding-on-fluent","task":"Spoken Language Understanding","dataset":"Fluent Speech Commands","model":"Pooling classifier pre-trained using force-aligned phoneme and word labels on LibriSpeech","rank_in_archive_order":15,"of":17,"metrics":{"Accuracy (%)":"98.8"},"uses_additional_data":true}],"syntology":{"syntology_url":"https://syntology.ai/paper/1904.03670","atlas_url":"https://app.syntology.ai/?focus=1904.03670","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}