{"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-voice-2-multi-speaker-neural-text-to","title":"Deep Voice 2: Multi-Speaker Neural Text-to-Speech","arxiv_id":"1705.08947","date":"2017-05-24","proceeding":"NeurIPS 2017 12","authors":["Sercan Arik","Gregory Diamos","Andrew Gibiansky","John Miller","Kainan Peng","Wei Ping","Jonathan Raiman","Yanqi Zhou"],"abstract":"We introduce a technique for augmenting neural text-to-speech (TTS) with\nlowdimensional trainable speaker embeddings to generate different voices from a\nsingle model. As a starting point, we show improvements over the two\nstate-ofthe-art approaches for single-speaker neural TTS: Deep Voice 1 and\nTacotron. We introduce Deep Voice 2, which is based on a similar pipeline with\nDeep Voice 1, but constructed with higher performance building blocks and\ndemonstrates a significant audio quality improvement over Deep Voice 1. We\nimprove Tacotron by introducing a post-processing neural vocoder, and\ndemonstrate a significant audio quality improvement. We then demonstrate our\ntechnique for multi-speaker speech synthesis for both Deep Voice 2 and Tacotron\non two multi-speaker TTS datasets. We show that a single neural TTS system can\nlearn hundreds of unique voices from less than half an hour of data per\nspeaker, while achieving high audio quality synthesis and preserving the\nspeaker identities almost perfectly.","url_abs":"http://arxiv.org/abs/1705.08947v2","url_pdf":"http://arxiv.org/pdf/1705.08947v2.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-voice-2-multi-speaker-neural-text-to","repo_url":"https://github.com/barronalex/Tacotron","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"tf","reach":null}],"tasks":[{"task_slug":"speech-synthesis","task_name":"Speech Synthesis"},{"task_slug":"text-to-speech","task_name":"Text to Speech"},{"task_slug":"text-to-speech-1","task_name":"text-to-speech"}],"methods":[{"method_slug":"batch-normalization","method_name":"Batch Normalization"},{"method_slug":"bigru","method_name":"BiGRU"},{"method_slug":"cbhg","method_name":"CBHG"},{"method_slug":"convolution","method_name":"Convolution"},{"method_slug":"dense-connections","method_name":"Dense Connections"},{"method_slug":"dropout","method_name":"Dropout"},{"method_slug":"gru","method_name":"GRU"},{"method_slug":"griffin-lim-algorithm","method_name":"Griffin-Lim Algorithm"},{"method_slug":"highway-layer","method_name":"Highway Layer"},{"method_slug":"highway-network","method_name":"Highway Network"},{"method_slug":"max-pooling","method_name":"Max Pooling"},{"method_slug":"relu","method_name":"ReLU"},{"method_slug":"residual-connection","method_name":"Residual Connection"},{"method_slug":"residual-gru","method_name":"Residual GRU"},{"method_slug":"sigmoid-activation","method_name":"Sigmoid Activation"},{"method_slug":"tacotron","method_name":"Tacotron"},{"method_slug":"tanh-activation","method_name":"Tanh Activation"}],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"syntology_url":null,"atlas_url":"https://app.syntology.ai/?focus=1705.08947","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}