[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"project-70761":3},{"id":4,"name":5,"fullName":6,"owner":7,"repo":5,"description":8,"homepage":9,"htmlUrl":10,"language":11,"languages":10,"totalLinesOfCode":10,"stars":12,"forks":13,"watchers":14,"openIssues":15,"contributorsCount":16,"subscribersCount":16,"size":16,"stars1d":17,"stars7d":18,"stars30d":19,"stars90d":16,"forks30d":16,"starsTrendScore":20,"compositeScore":21,"rankGlobal":10,"rankLanguage":10,"license":22,"archived":23,"fork":23,"defaultBranch":24,"hasWiki":25,"hasPages":23,"topics":26,"createdAt":10,"pushedAt":10,"updatedAt":47,"readmeContent":48,"aiSummary":49,"trendingCount":16,"starSnapshotCount":16,"syncStatus":50,"lastSyncTime":51,"discoverSource":52},70761,"PaddleSpeech","PaddlePaddle\u002FPaddleSpeech","PaddlePaddle","Easy-to-use Speech Toolkit including Self-Supervised Learning model, SOTA\u002FStreaming ASR with punctuation, Streaming TTS with text frontend, Speaker Verification System, End-to-End Speech Translation and Keyword Spotting. Won NAACL2022 Best Demo Award.","https:\u002F\u002Fpaddlespeech.readthedocs.io",null,"Python",12614,1957,187,266,0,1,4,22,3,76.08,"Apache License 2.0",false,"develop",true,[27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46],"asr","code-switch","conformer","kws","punctuation-restoration","self-supervised-learning","sound-classification","speech-alignment","speech-recognition","speech-synthesis","speech-translation","streaming-asr","streaming-tts","transformer","tts","vocoder","voice-cloning","voice-recognition","wav2vec2","whisper","2026-06-12 04:00:57","([简体中文](.\u002FREADME_cn.md)|English)\n\u003Cp align=\"center\">\n  \u003Cimg src=\".\u002Fdocs\u002Fimages\u002FPaddleSpeech_logo.png\" \u002F>\n\u003C\u002Fp>\n\n\u003Cp align=\"center\">\n    \u003Ca href=\".\u002FLICENSE\">\u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fbadge\u002Flicense-Apache%202-red.svg\">\u003C\u002Fa>\n    \u003Ca href=\"https:\u002F\u002Fgithub.com\u002FPaddlePaddle\u002FPaddleSpeech\u002Freleases\">\u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fv\u002Frelease\u002FPaddlePaddle\u002FPaddleSpeech?color=ffa\">\u003C\u002Fa>\n    \u003Ca href=\"support os\">\u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fbadge\u002Fos-linux%2C%20win%2C%20mac-pink.svg\">\u003C\u002Fa>\n    \u003Ca href=\"\">\u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fbadge\u002Fpython-3.8+-aff.svg\">\u003C\u002Fa>\n    \u003Ca href=\"https:\u002F\u002Fgithub.com\u002FPaddlePaddle\u002FPaddleSpeech\u002Fgraphs\u002Fcontributors\">\u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fcontributors\u002FPaddlePaddle\u002FPaddleSpeech?color=9ea\">\u003C\u002Fa>\n    \u003Ca href=\"https:\u002F\u002Fgithub.com\u002FPaddlePaddle\u002FPaddleSpeech\u002Fcommits\">\u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fcommit-activity\u002Fm\u002FPaddlePaddle\u002FPaddleSpeech?color=3af\">\u003C\u002Fa>\n    \u003Ca href=\"https:\u002F\u002Fgithub.com\u002FPaddlePaddle\u002FPaddleSpeech\u002Fissues\">\u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fissues\u002FPaddlePaddle\u002FPaddleSpeech?color=9cc\">\u003C\u002Fa>\n    \u003Ca href=\"https:\u002F\u002Fgithub.com\u002FPaddlePaddle\u002FPaddleSpeech\u002Fstargazers\">\u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002FPaddlePaddle\u002FPaddleSpeech?color=ccf\">\u003C\u002Fa>\n    \u003Ca href=\"=https:\u002F\u002Fpypi.org\u002Fproject\u002Fpaddlespeech\u002F\">\u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fpypi\u002Fdm\u002FPaddleSpeech\">\u003C\u002Fa>\n    \u003Ca href=\"=https:\u002F\u002Fpypi.org\u002Fproject\u002Fpaddlespeech\u002F\">\u003Cimg src=\"https:\u002F\u002Fstatic.pepy.tech\u002Fbadge\u002Fpaddlespeech\">\u003C\u002Fa>\n    \u003Ca href=\"https:\u002F\u002Fhuggingface.co\u002Fspaces\">\u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fbadge\u002F%F0%9F%A4%97%20Hugging%20Face-Spaces-blue\">\u003C\u002Fa>\n\u003C\u002Fp>\n\u003Cdiv align=\"center\">  \n\u003Ch4>\n    \u003Ca href=\"#quick-start\"> Quick Start \u003C\u002Fa>\n  | \u003Ca href=\"#documents\"> Documents \u003C\u002Fa>\n  | \u003Ca href=\"#model-list\"> Models List \u003C\u002Fa>\n  | \u003Ca href=\"https:\u002F\u002Faistudio.baidu.com\u002Faistudio\u002Fcourse\u002Fintroduce\u002F25130\"> AIStudio Courses \u003C\u002Fa>\n  | \u003Ca href=\"https:\u002F\u002Farxiv.org\u002Fabs\u002F2205.12007\"> NAACL2022 Best Demo Award Paper \u003C\u002Fa>\n  | \u003Ca href=\"https:\u002F\u002Fgitee.com\u002Fpaddlepaddle\u002FPaddleSpeech\"> Gitee \u003C\u002Fa>\n\u003C\u002Fh4>\n\u003C\u002Fdiv>\n\n------------------------------------------------------------------------------------\n\n**PaddleSpeech** is an open-source toolkit on [PaddlePaddle](https:\u002F\u002Fgithub.com\u002FPaddlePaddle\u002FPaddle) platform for a variety of critical tasks in speech and audio, with the state-of-art and influential models. \n\n**PaddleSpeech** won the [NAACL2022 Best Demo Award](https:\u002F\u002F2022.naacl.org\u002Fblog\u002Fbest-demo-award\u002F), please check out our paper on [Arxiv](https:\u002F\u002Farxiv.org\u002Fabs\u002F2205.12007).\n\n##### Speech Recognition\n\n\u003Cdiv align = \"center\">\n\u003Ctable style=\"width:100%\">\n  \u003Cthead>\n    \u003Ctr>\n      \u003Cth> Input Audio  \u003C\u002Fth>\n      \u003Cth width=\"550\"> Recognition Result  \u003C\u002Fth>\n    \u003C\u002Ftr>\n  \u003C\u002Fthead>\n  \u003Ctbody>\n   \u003Ctr>\n      \u003Ctd align = \"center\">\n      \u003Ca href=\"https:\u002F\u002Fpaddlespeech.cdn.bcebos.com\u002FPaddleAudio\u002Fen.wav\" rel=\"nofollow\">\n            \u003Cimg align=\"center\" src=\".\u002Fdocs\u002Fimages\u002Faudio_icon.png\" width=\"200 style=\"max-width: 100%;\">\u003C\u002Fa>\u003Cbr>\n      \u003C\u002Ftd>\n      \u003Ctd >I knocked at the door on the ancient side of the building.\u003C\u002Ftd>\n    \u003C\u002Ftr>\n    \u003Ctr>\n      \u003Ctd align = \"center\">\n      \u003Ca href=\"https:\u002F\u002Fpaddlespeech.cdn.bcebos.com\u002FPaddleAudio\u002Fzh.wav\" rel=\"nofollow\">\n            \u003Cimg align=\"center\" src=\".\u002Fdocs\u002Fimages\u002Faudio_icon.png\" width=\"200\" style=\"max-width: 100%;\">\u003C\u002Fa>\u003Cbr>\n      \u003C\u002Ftd>\n      \u003Ctd>我认为跑步最重要的就是给我带来了身体健康。\u003C\u002Ftd>\n    \u003C\u002Ftr>\n  \u003C\u002Ftbody>\n\u003C\u002Ftable>\n\n\u003C\u002Fdiv>\n\n##### Speech Translation (English to Chinese)\n\n\u003Cdiv align = \"center\">\n\u003Ctable style=\"width:100%\">\n  \u003Cthead>\n    \u003Ctr>\n      \u003Cth> Input Audio  \u003C\u002Fth>\n      \u003Cth width=\"550\"> Translations Result  \u003C\u002Fth>\n    \u003C\u002Ftr>\n  \u003C\u002Fthead>\n  \u003Ctbody>\n   \u003Ctr>\n      \u003Ctd align = \"center\">\n      \u003Ca href=\"https:\u002F\u002Fpaddlespeech.cdn.bcebos.com\u002FPaddleAudio\u002Fen.wav\" rel=\"nofollow\">\n            \u003Cimg align=\"center\" src=\".\u002Fdocs\u002Fimages\u002Faudio_icon.png\" width=\"200 style=\"max-width: 100%;\">\u003C\u002Fa>\u003Cbr>\n      \u003C\u002Ftd>\n      \u003Ctd >我 在 这栋 建筑 的 古老 门上 敲门。\u003C\u002Ftd>\n    \u003C\u002Ftr>\n  \u003C\u002Ftbody>\n\u003C\u002Ftable>\n\n\u003C\u002Fdiv>\n\n##### Text-to-Speech\n\u003Cdiv align = \"center\">\n\u003Ctable style=\"width:100%\">\n  \u003Cthead>\n    \u003Ctr>\n      \u003Cth width=\"550\" > Input Text\u003C\u002Fth>\n      \u003Cth>Synthetic Audio\u003C\u002Fth>\n    \u003C\u002Ftr>\n  \u003C\u002Fthead>\n  \u003Ctbody>\n   \u003Ctr>\n      \u003Ctd>Life was like a box of chocolates, you never know what you're gonna get.\u003C\u002Ftd>\n      \u003Ctd align = \"center\">\n      \u003Ca href=\"https:\u002F\u002Fpaddlespeech.cdn.bcebos.com\u002FParakeet\u002Fdocs\u002Fdemos\u002Ftacotron2_ljspeech_waveflow_samples_0.2\u002Fsentence_1.wav\" rel=\"nofollow\">\n            \u003Cimg align=\"center\" src=\".\u002Fdocs\u002Fimages\u002Faudio_icon.png\" width=\"200\" style=\"max-width: 100%;\">\u003C\u002Fa>\u003Cbr>\n      \u003C\u002Ftd>\n    \u003C\u002Ftr>\n    \u003Ctr>\n      \u003Ctd>早上好，今天是2020\u002F10\u002F29，最低温度是-3°C。\u003C\u002Ftd>\n      \u003Ctd align = \"center\">\n      \u003Ca href=\"https:\u002F\u002Fpaddlespeech.cdn.bcebos.com\u002FParakeet\u002Fdocs\u002Fdemos\u002Fparakeet_espnet_fs2_pwg_demo\u002Ftn_g2p\u002Fparakeet\u002F001.wav\" rel=\"nofollow\">\n            \u003Cimg align=\"center\" src=\".\u002Fdocs\u002Fimages\u002Faudio_icon.png\" width=\"200\" style=\"max-width: 100%;\">\u003C\u002Fa>\u003Cbr>\n      \u003C\u002Ftd>\n    \u003C\u002Ftr>\n    \u003Ctr>\n      \u003Ctd>季姬寂，集鸡，鸡即棘鸡。棘鸡饥叽，季姬及箕稷济鸡。鸡既济，跻姬笈，季姬忌，急咭鸡，鸡急，继圾几，季姬急，即籍箕击鸡，箕疾击几伎，伎即齑，鸡叽集几基，季姬急极屐击鸡，鸡既殛，季姬激，即记《季姬击鸡记》。\u003C\u002Ftd>\n      \u003Ctd align = \"center\">\n      \u003Ca href=\"https:\u002F\u002Fpaddlespeech.cdn.bcebos.com\u002FParakeet\u002Fdocs\u002Fdemos\u002Fjijiji.wav\" rel=\"nofollow\">\n            \u003Cimg align=\"center\" src=\".\u002Fdocs\u002Fimages\u002Faudio_icon.png\" width=\"200\" style=\"max-width: 100%;\">\u003C\u002Fa>\u003Cbr>\n      \u003C\u002Ftd>\n    \u003C\u002Ftr>\n    \u003Ctr>\n      \u003Ctd>大家好，我是 parrot 虚拟老师，我们来读一首诗，我与春风皆过客，I and the spring breeze are passing by，你携秋水揽星河，you take the autumn water to take the galaxy。\u003C\u002Ftd>\n      \u003Ctd align = \"center\">\n      \u003Ca href=\"https:\u002F\u002Fpaddlespeech.cdn.bcebos.com\u002FParakeet\u002Fdocs\u002Fdemos\u002Flabixiaoxin.wav\" rel=\"nofollow\">\n            \u003Cimg align=\"center\" src=\".\u002Fdocs\u002Fimages\u002Faudio_icon.png\" width=\"200\" style=\"max-width: 100%;\">\u003C\u002Fa>\u003Cbr>\n      \u003C\u002Ftd>\n    \u003C\u002Ftr>\n    \u003Ctr>\n      \u003Ctd>宜家唔系事必要你讲，但系你所讲嘅说话将会变成呈堂证供。\u003C\u002Ftd>\n      \u003Ctd align = \"center\">\n      \u003Ca href=\"https:\u002F\u002Fpaddlespeech.cdn.bcebos.com\u002FParakeet\u002Fdocs\u002Fdemos\u002Fchengtangzhenggong.wav\" rel=\"nofollow\">\n            \u003Cimg align=\"center\" src=\".\u002Fdocs\u002Fimages\u002Faudio_icon.png\" width=\"200\" style=\"max-width: 100%;\">\u003C\u002Fa>\u003Cbr>\n      \u003C\u002Ftd>\n    \u003C\u002Ftr>\n    \u003Ctr>\n      \u003Ctd>各个国家有各个国家嘅国歌\u003C\u002Ftd>\n      \u003Ctd align = \"center\">\n      \u003Ca href=\"https:\u002F\u002Fpaddlespeech.cdn.bcebos.com\u002FParakeet\u002Fdocs\u002Fdemos\u002Fgegege.wav\" rel=\"nofollow\">\n            \u003Cimg align=\"center\" src=\".\u002Fdocs\u002Fimages\u002Faudio_icon.png\" width=\"200\" style=\"max-width: 100%;\">\u003C\u002Fa>\u003Cbr>\n      \u003C\u002Ftd>\n    \u003C\u002Ftr>\n  \u003C\u002Ftbody>\n\u003C\u002Ftable>\n\n\u003C\u002Fdiv>\n\nFor more synthesized audios, please refer to [PaddleSpeech Text-to-Speech samples](https:\u002F\u002Fpaddlespeech.readthedocs.io\u002Fen\u002Flatest\u002Ftts\u002Fdemo.html).\n\n##### Punctuation Restoration\n\u003Cdiv align = \"center\">\n\u003Ctable style=\"width:100%\">\n  \u003Cthead>\n    \u003Ctr>\n      \u003Cth width=\"390\"> Input Text \u003C\u002Fth>\n      \u003Cth width=\"390\"> Output Text \u003C\u002Fth>\n    \u003C\u002Ftr>\n  \u003C\u002Fthead>\n  \u003Ctbody>\n   \u003Ctr>\n      \u003Ctd>今天的天气真不错啊你下午有空吗我想约你一起去吃饭\u003C\u002Ftd>\n      \u003Ctd>今天的天气真不错啊！你下午有空吗？我想约你一起去吃饭。\u003C\u002Ftd>\n    \u003C\u002Ftr>\n  \u003C\u002Ftbody>\n\u003C\u002Ftable>\n\n\u003C\u002Fdiv>\n\n\n### Features\n\nVia the easy-to-use, efficient, flexible and scalable implementation, our vision is to empower both industrial application and academic research, including training, inference & testing modules, and deployment process. To be more specific, this toolkit features at:\n- 📦  **Ease of Use**: low barriers to install, [CLI](#quick-start), [Server](#quick-start-server), and [Streaming Server](#quick-start-streaming-server) is available to quick-start your journey.\n- 🏆  **Align to the State-of-the-Art**: we provide high-speed and ultra-lightweight models, and also cutting-edge technology. \n- 🏆  **Streaming ASR and TTS System**: we provide production ready streaming asr and streaming tts system.\n- 💯  **Rule-based Chinese frontend**: our frontend contains Text Normalization and Grapheme-to-Phoneme (G2P, including Polyphone and Tone Sandhi). Moreover, we use self-defined linguistic rules to adapt Chinese context.\n- 📦  **Varieties of Functions that Vitalize both Industrial and Academia**:\n  - 🛎️  *Implementation of critical audio tasks*: this toolkit contains audio functions like  Automatic Speech Recognition, Text-to-Speech Synthesis, Speaker Verification, KeyWord Spotting, Audio Classification, and Speech Translation, etc.\n  - 🔬  *Integration of mainstream models and datasets*: the toolkit implements modules that participate in the whole pipeline of the speech tasks, and uses mainstream datasets like LibriSpeech, LJSpeech, AIShell, CSMSC, etc. See also [model list](#model-list) for more details.\n  - 🧩  *Cascaded models application*: as an extension of the typical traditional audio tasks, we combine the workflows of the aforementioned tasks with other fields like Natural language processing (NLP) and Computer Vision (CV).\n\n### Recent Update\n- 🎉 2025.09.01: Add [Whisper large v3 and turbo model](https:\u002F\u002Fgithub.com\u002FPaddlePaddle\u002FPaddleSpeech\u002Ftree\u002Fdevelop\u002Fdemos\u002Fwhisper).\n- 🤗 2025.08.11: Add [code-switch online model and server demo](.\u002Fexamples\u002Ftal_cs\u002Fasr1\u002F).\n- 👑 2023.05.31: Add [WavLM ASR-en](https:\u002F\u002Fgithub.com\u002FPaddlePaddle\u002FPaddleSpeech\u002Fblob\u002Fdevelop\u002Fexamples\u002Flibrispeech\u002Fasr5), WavLM fine-tuning for ASR on LibriSpeech.\n- 🎉 2023.05.18: Add [Squeezeformer](https:\u002F\u002Fgithub.com\u002FPaddlePaddle\u002FPaddleSpeech\u002Ftree\u002Fdevelop\u002Fexamples\u002Faishell\u002Fasr1), Squeezeformer training for ASR on Aishell.\n- 👑 2023.05.04: Add [HuBERT ASR-en](https:\u002F\u002Fgithub.com\u002FPaddlePaddle\u002FPaddleSpeech\u002Fblob\u002Fdevelop\u002Fexamples\u002Flibrispeech\u002Fasr4), HuBERT fine-tuning for ASR on LibriSpeech.\n- ⚡ 2023.04.28: Fix [0-d tensor](https:\u002F\u002Fgithub.com\u002FPaddlePaddle\u002FPaddleSpeech\u002Fpull\u002F3214), with the upgrade of paddlepaddle==2.5, the problem of modifying 0-d tensor has been solved.\n- 👑 2023.04.25: Add [AMP for U2 conformer](https:\u002F\u002Fgithub.com\u002FPaddlePaddle\u002FPaddleSpeech\u002Fpull\u002F3167).\n- 🔥 2023.04.06: Add [subtitle file (.srt format) generation example](.\u002Fdemos\u002Fstreaming_asr_server).\n- 🔥 2023.03.14: Add SVS(Singing Voice Synthesis) examples with Opencpop dataset, including [DiffSinger](.\u002Fexamples\u002Fopencpop\u002Fsvs1)、[PWGAN](.\u002Fexamples\u002Fopencpop\u002Fvoc1) and [HiFiGAN](.\u002Fexamples\u002Fopencpop\u002Fvoc5), the effect is continuously optimized.\n- 👑 2023.03.09: Add [Wav2vec2ASR-zh](.\u002Fexamples\u002Faishell\u002Fasr3).\n- 🎉 2023.03.07: Add [TTS ARM Linux C++ Demo (with C++ Chinese Text Frontend)](.\u002Fdemos\u002FTTSArmLinux).\n- 🔥 2023.03.03 Add Voice Conversion [StarGANv2-VC synthesize pipeline](.\u002Fexamples\u002Fvctk\u002Fvc3).\n- 🎉 2023.02.16: Add [Cantonese TTS](.\u002Fexamples\u002Fcanton\u002Ftts3).\n- 🔥 2023.01.10: Add [code-switch asr CLI and Demos](.\u002Fdemos\u002Fspeech_recognition).\n- 👑 2023.01.06: Add [code-switch asr tal_cs recipe](.\u002Fexamples\u002Ftal_cs\u002Fasr1\u002F).\n- 🎉 2022.12.02: Add [end-to-end Prosody Prediction pipeline](.\u002Fexamples\u002Fcsmsc\u002Ftts3_rhy) (including using prosody labels in Acoustic Model).\n- 🎉 2022.11.30: Add [TTS Android Demo](.\u002Fdemos\u002FTTSAndroid).\n- 🤗 2022.11.28: PP-TTS and PP-ASR demos are available in [AIStudio](https:\u002F\u002Faistudio.baidu.com\u002Faistudio\u002Fmodelsoverview) and [official website\n of paddlepaddle](https:\u002F\u002Fwww.paddlepaddle.org.cn\u002Fmodels).\n- 👑 2022.11.18: Add [Whisper CLI and Demos](https:\u002F\u002Fgithub.com\u002FPaddlePaddle\u002FPaddleSpeech\u002Fpull\u002F2640), support multi language recognition and translation.\n- 🔥 2022.11.18: Add [Wav2vec2 CLI and Demos](.\u002Fdemos\u002Fspeech_ssl), Support ASR and Feature Extraction.\n- 🎉 2022.11.17: Add [male voice for TTS](https:\u002F\u002Fgithub.com\u002FPaddlePaddle\u002FPaddleSpeech\u002Fpull\u002F2660).\n- 🔥 2022.11.07: Add [U2\u002FU2++ C++ High Performance Streaming ASR Deployment](https:\u002F\u002Fgithub.com\u002FPaddlePaddle\u002FPaddleSpeech\u002Fblob\u002Fdevelop\u002Fruntime\u002Fexamples\u002Fu2pp_ol\u002Fwenetspeech).\n- 👑 2022.11.01: Add [Adversarial Loss](https:\u002F\u002Farxiv.org\u002Fpdf\u002F1907.04448.pdf) for [Chinese English mixed TTS](.\u002Fexamples\u002Fzh_en_tts\u002Ftts3).\n- 🔥 2022.10.26: Add [Prosody Prediction](.\u002Fexamples\u002Fother\u002Frhy) for TTS.\n- 🎉 2022.10.21: Add [SSML](https:\u002F\u002Fgithub.com\u002FPaddlePaddle\u002FPaddleSpeech\u002Fdiscussions\u002F2538) for TTS Chinese Text Frontend.\n- 👑 2022.10.11: Add [Wav2vec2ASR-en](.\u002Fexamples\u002Flibrispeech\u002Fasr3), wav2vec2.0 fine-tuning for ASR on LibriSpeech.\n- 🔥 2022.09.26: Add Voice Cloning, TTS finetune, and [ERNIE-SAT](https:\u002F\u002Farxiv.org\u002Fabs\u002F2211.03545) in [PaddleSpeech Web Demo](.\u002Fdemos\u002Fspeech_web).\n- ⚡ 2022.09.09: Add AISHELL-3 Voice Cloning [example](.\u002Fexamples\u002Faishell3\u002Fvc2) with ECAPA-TDNN speaker encoder.\n- ⚡ 2022.08.25: Release TTS [finetune](.\u002Fexamples\u002Fother\u002Ftts_finetune\u002Ftts3) example.\n- 🔥 2022.08.22: Add [ERNIE-SAT](https:\u002F\u002Farxiv.org\u002Fabs\u002F2211.03545) models: [ERNIE-SAT-vctk](.\u002Fexamples\u002Fvctk\u002Fernie_sat)、[ERNIE-SAT-aishell3](.\u002Fexamples\u002Faishell3\u002Fernie_sat)、[ERNIE-SAT-zh_en](.\u002Fexamples\u002Faishell3_vctk\u002Fernie_sat).\n- 🔥 2022.08.15: Add [g2pW](https:\u002F\u002Fgithub.com\u002FGitYCC\u002Fg2pW) into TTS Chinese Text Frontend.\n- 🔥 2022.08.09: Release [Chinese English mixed TTS](.\u002Fexamples\u002Fzh_en_tts\u002Ftts3).\n- ⚡ 2022.08.03: Add ONNXRuntime infer for  TTS CLI.\n- 🎉 2022.07.18: Release VITS: [VITS-csmsc](.\u002Fexamples\u002Fcsmsc\u002Fvits)、[VITS-aishell3](.\u002Fexamples\u002Faishell3\u002Fvits)、[VITS-VC](.\u002Fexamples\u002Faishell3\u002Fvits-vc).\n- 🎉 2022.06.22: All TTS models support ONNX format.\n- 🍀 2022.06.17: Add [PaddleSpeech Web Demo](.\u002Fdemos\u002Fspeech_web).\n- 👑 2022.05.13: Release [PP-ASR](.\u002Fdocs\u002Fsource\u002Fasr\u002FPPASR.md)、[PP-TTS](.\u002Fdocs\u002Fsource\u002Ftts\u002FPPTTS.md)、[PP-VPR](docs\u002Fsource\u002Fvpr\u002FPPVPR.md).\n- 👏🏻 2022.05.06: `PaddleSpeech Streaming Server` is available for `Streaming ASR` with `Punctuation Restoration` and `Token Timestamp` and `Text-to-Speech`.\n- 👏🏻 2022.05.06: `PaddleSpeech Server` is available for `Audio Classification`, `Automatic Speech Recognition` and `Text-to-Speech`, `Speaker Verification` and `Punctuation Restoration`.\n- 👏🏻 2022.03.28: `PaddleSpeech CLI` is available for `Speaker Verification`.\n- 👏🏻 2021.12.10: `PaddleSpeech CLI` is available for `Audio Classification`, `Automatic Speech Recognition`, `Speech Translation (English to Chinese)` and `Text-to-Speech`.\n\n### Community\n- Scan the QR code below with your Wechat, you can access to official technical exchange group and get the bonus ( more than 20GB learning materials, such as papers, codes and videos ) and the live link of the lessons. Look forward to your participation.\n\n\u003Cdiv align=\"center\">\n\u003Cimg src=\"https:\u002F\u002Fuser-images.githubusercontent.com\u002F30135920\u002F212860467-9e943cc3-8be8-49a4-97fd-7c94aad8e979.jpg\"  width = \"200\"  \u002F>\n\u003C\u002Fdiv>\n\n## Installation\n\nWe strongly recommend our users to install PaddleSpeech in **Linux** with *python>=3.8*. \n\n### **Dependency Introduction**\n\n+ gcc >= 4.8.5\n+ paddlepaddle\n+ python >= 3.8\n+ OS support:  Linux(recommend), Windows, Mac OSX\n\nPaddleSpeech depends on paddlepaddle. For installation, please refer to the official website of [paddlepaddle](https:\u002F\u002Fwww.paddlepaddle.org.cn\u002Fen) and choose according to your own machine. Here is an example of the cpu version.\n\n```bash\npip install paddlepaddle -i https:\u002F\u002Fmirror.baidu.com\u002Fpypi\u002Fsimple\n```\nYou can also specify the version of paddlepaddle or install the develop version. \n```bash\n# install 2.4.1 version. Note, 2.4.1 is just an example, please follow the minimum dependency of paddlepaddle for your selection\npip install paddlepaddle==2.4.1 -i https:\u002F\u002Fmirror.baidu.com\u002Fpypi\u002Fsimple\n# install develop version\npip install paddlepaddle==0.0.0 -f https:\u002F\u002Fwww.paddlepaddle.org.cn\u002Fwhl\u002Flinux\u002Fcpu-mkl\u002Fdevelop.html\n```\n\nThere are two quick installation methods for PaddleSpeech, one is pip installation, and the other is source code compilation (recommended).\n### pip install\n\n```shell\npip install pytest-runner\npip install paddlespeech\n```\n\n### source code compilation\n\n```shell\ngit clone https:\u002F\u002Fgithub.com\u002FPaddlePaddle\u002FPaddleSpeech.git\ncd PaddleSpeech\npip install pytest-runner\npip install .\n# If you need to install in editable mode, you need to use --use-pep517. The command is as follows:\n# pip install -e . --use-pep517\n```\n\nFor more installation problems, such as conda environment, librosa-dependent, gcc problems, kaldi installation, etc., you can refer to this [installation document](.\u002Fdocs\u002Fsource\u002Finstall.md). If you encounter problems during installation, you can leave a message on [#2150](https:\u002F\u002Fgithub.com\u002FPaddlePaddle\u002FPaddleSpeech\u002Fissues\u002F2150) and find related problems\n\n\n\u003Ca name=\"quickstart\">\u003C\u002Fa>\n## Quick Start\n\nDevelopers can have a try of our models with [PaddleSpeech Command Line](.\u002Fpaddlespeech\u002Fcli\u002FREADME.md) or Python. Change `--input` to test your own audio\u002Ftext and support 16k wav format audio.\n\n**You can also quickly experience it in AI Studio 👉🏻 [PaddleSpeech API Demo](https:\u002F\u002Faistudio.baidu.com\u002Faistudio\u002Fprojectdetail\u002F4353348?sUid=2470186&shared=1&ts=1660876445786)**\n\n\nTest audio sample download\n\n```shell\nwget -c https:\u002F\u002Fpaddlespeech.cdn.bcebos.com\u002FPaddleAudio\u002Fzh.wav\nwget -c https:\u002F\u002Fpaddlespeech.cdn.bcebos.com\u002FPaddleAudio\u002Fen.wav\n```\n\n### Automatic Speech Recognition\n\n\u003Cdetails>\u003Csummary> （Click to expand）Open Source Speech Recognition\u003C\u002Fsummary>\n\n**command line experience**\n\n```shell\npaddlespeech asr --lang zh --input zh.wav\n```\n\n**Python API experience**\n\n```python\n>>> from paddlespeech.cli.asr.infer import ASRExecutor\n>>> asr = ASRExecutor()\n>>> result = asr(audio_file=\"zh.wav\")\n>>> print(result)\n我认为跑步最重要的就是给我带来了身体健康\n```\n\u003C\u002Fdetails>\n\n### Text-to-Speech\n\n\u003Cdetails>\u003Csummary> Open Source Speech Synthesis\u003C\u002Fsummary>\n\nOutput 24k sample rate wav format audio\n\n\n**command line experience**\n\n```shell\npaddlespeech tts --input \"你好，欢迎使用百度飞桨深度学习框架！\" --output output.wav\n```\n\n**Python API experience**\n\n```python\n>>> from paddlespeech.cli.tts.infer import TTSExecutor\n>>> tts = TTSExecutor()\n>>> tts(text=\"今天天气十分不错。\", output=\"output.wav\")\n```\n- You can experience in [Huggingface Spaces](https:\u002F\u002Fhuggingface.co\u002Fspaces) [TTS Demo](https:\u002F\u002Fhuggingface.co\u002Fspaces\u002FKPatrick\u002FPaddleSpeechTTS)\n\n\u003C\u002Fdetails>\n\n### Audio Classification\n\n\u003Cdetails>\u003Csummary> An open-domain sound classification tool\u003C\u002Fsummary>\n\nSound classification model based on 527 categories of AudioSet dataset\n\n**command line experience**\n\n```shell\npaddlespeech cls --input zh.wav\n```\n\n**Python API experience**\n\n```python\n>>> from paddlespeech.cli.cls.infer import CLSExecutor\n>>> cls = CLSExecutor()\n>>> result = cls(audio_file=\"zh.wav\")\n>>> print(result)\nSpeech 0.9027186632156372\n```\n\n\u003C\u002Fdetails>\n\n### Voiceprint Extraction\n\n\u003Cdetails>\u003Csummary> Industrial-grade voiceprint extraction tool\u003C\u002Fsummary>\n\n**command line experience**\n\n```shell\npaddlespeech vector --task spk --input zh.wav\n```\n\n**Python API experience**\n\n```python\n>>> from paddlespeech.cli.vector import VectorExecutor\n>>> vec = VectorExecutor()\n>>> result = vec(audio_file=\"zh.wav\")\n>>> print(result) # 187维向量\n[ -0.19083306   9.474295   -14.122263    -2.0916545    0.04848729\n   4.9295826    1.4780062    0.3733844   10.695862     3.2697146\n  -4.48199     -0.6617882   -9.170393   -11.1568775   -1.2358263 ...]\n```\n\n\u003C\u002Fdetails>\n\n### Punctuation Restoration\n\n\u003Cdetails>\u003Csummary> Quick recovery of text punctuation, works with ASR models\u003C\u002Fsummary>\n\n**command line experience**\n\n```shell\npaddlespeech text --task punc --input 今天的天气真不错啊你下午有空吗我想约你一起去吃饭\n```\n\n**Python API experience**\n\n```python\n>>> from paddlespeech.cli.text.infer import TextExecutor\n>>> text_punc = TextExecutor()\n>>> result = text_punc(text=\"今天的天气真不错啊你下午有空吗我想约你一起去吃饭\")\n今天的天气真不错啊！你下午有空吗？我想约你一起去吃饭。\n```\n\n\u003C\u002Fdetails>\n\n### Speech Translation\n\n\u003Cdetails>\u003Csummary> End-to-end English to Chinese Speech Translation Tool\u003C\u002Fsummary>\n\nUse pre-compiled kaldi related tools, only support experience in Ubuntu system\n\n**command line experience**\n\n```shell\npaddlespeech st --input en.wav\n```\n\n**Python API experience**\n\n```python\n>>> from paddlespeech.cli.st.infer import STExecutor\n>>> st = STExecutor()\n>>> result = st(audio_file=\"en.wav\")\n['我 在 这栋 建筑 的 古老 门上 敲门 。']\n```\n\n\u003C\u002Fdetails>\n\n\n\u003Ca name=\"quickstartserver\">\u003C\u002Fa>\n## Quick Start Server\n\nDevelopers can have a try of our speech server with [PaddleSpeech Server Command Line](.\u002Fpaddlespeech\u002Fserver\u002FREADME.md).\n\n**You can try it quickly in AI Studio (recommend): [SpeechServer](https:\u002F\u002Faistudio.baidu.com\u002Faistudio\u002Fprojectdetail\u002F4354592?sUid=2470186&shared=1&ts=1660877827034)**\n\n**Start server**     \n\n```shell\npaddlespeech_server start --config_file .\u002Fdemos\u002Fspeech_server\u002Fconf\u002Fapplication.yaml\n```\n\n**Access Speech Recognition Services**     \n\n```shell\npaddlespeech_client asr --server_ip 127.0.0.1 --port 8090 --input input_16k.wav\n```\n\n**Access Text to Speech Services**     \n\n```shell\npaddlespeech_client tts --server_ip 127.0.0.1 --port 8090 --input \"您好，欢迎使用百度飞桨语音合成服务。\" --output output.wav\n```\n\n**Access Audio Classification Services**     \n```shell\npaddlespeech_client cls --server_ip 127.0.0.1 --port 8090 --input input.wav\n```\n\n\nFor more information about server command lines, please see: [speech server demos](https:\u002F\u002Fgithub.com\u002FPaddlePaddle\u002FPaddleSpeech\u002Ftree\u002Fdevelop\u002Fdemos\u002Fspeech_server)\n\n\n\u003Ca name=\"quickstartstreamingserver\">\u003C\u002Fa>\n## Quick Start Streaming Server\n\nDevelopers can have a try of  [streaming asr](.\u002Fdemos\u002Fstreaming_asr_server\u002FREADME.md) and [streaming tts](.\u002Fdemos\u002Fstreaming_tts_server\u002FREADME.md) server.\n\n**Start Streaming Speech Recognition Server**\n\n```\npaddlespeech_server start --config_file .\u002Fdemos\u002Fstreaming_asr_server\u002Fconf\u002Fapplication.yaml\n```\n\n**Access Streaming Speech Recognition Services**     \n\n```\npaddlespeech_client asr_online --server_ip 127.0.0.1 --port 8090 --input input_16k.wav\n```\n\n**Start Streaming Text to Speech  Server**\n\n```\npaddlespeech_server start --config_file .\u002Fdemos\u002Fstreaming_tts_server\u002Fconf\u002Ftts_online_application.yaml\n```\n\n**Access Streaming Text to Speech Services**     \n\n```\npaddlespeech_client tts_online --server_ip 127.0.0.1 --port 8092 --protocol http --input \"您好，欢迎使用百度飞桨语音合成服务。\" --output output.wav\n```\n\nFor more information please see:  [streaming asr](.\u002Fdemos\u002Fstreaming_asr_server\u002FREADME.md) and [streaming tts](.\u002Fdemos\u002Fstreaming_tts_server\u002FREADME.md) \n\n\u003Ca name=\"ModelList\">\u003C\u002Fa>\n\n## Model List\n\nPaddleSpeech supports a series of most popular models. They are summarized in [released models](.\u002Fdocs\u002Fsource\u002Freleased_model.md) and attached with available pretrained models.\n\n\u003Ca name=\"SpeechToText\">\u003C\u002Fa>\n\n**Speech-to-Text** contains *Acoustic Model*, *Language Model*, and *Speech Translation*, with the following details:\n\n\u003Ctable style=\"width:100%\">\n  \u003Cthead>\n    \u003Ctr>\n      \u003Cth>Speech-to-Text Module Type\u003C\u002Fth>\n      \u003Cth>Dataset\u003C\u002Fth>\n      \u003Cth>Model Type\u003C\u002Fth>\n      \u003Cth>Example\u003C\u002Fth>\n    \u003C\u002Ftr>\n  \u003C\u002Fthead>\n  \u003Ctbody>\n    \u003Ctr>\n      \u003Ctd rowspan=\"4\">Speech Recogination\u003C\u002Ftd>\n      \u003Ctd rowspan=\"2\" >Aishell\u003C\u002Ftd>\n      \u003Ctd >DeepSpeech2 RNN + Conv based Models\u003C\u002Ftd>\n      \u003Ctd>\n      \u003Ca href = \".\u002Fexamples\u002Faishell\u002Fasr0\">deepspeech2-aishell\u003C\u002Fa>\n      \u003C\u002Ftd>\n    \u003C\u002Ftr>\n    \u003Ctr>\n      \u003Ctd>Transformer based Attention Models \u003C\u002Ftd>\n      \u003Ctd>\n      \u003Ca href = \".\u002Fexamples\u002Faishell\u002Fasr1\">u2.transformer.conformer-aishell\u003C\u002Fa>\n      \u003C\u002Ftd>\n    \u003C\u002Ftr>\n    \u003Ctr>\n      \u003Ctd> Librispeech\u003C\u002Ftd>\n      \u003Ctd>Transformer based Attention Models \u003C\u002Ftd>\n      \u003Ctd>\n      \u003Ca href = \".\u002Fexamples\u002Flibrispeech\u002Fasr0\">deepspeech2-librispeech\u003C\u002Fa> \u002F \u003Ca href = \".\u002Fexamples\u002Flibrispeech\u002Fasr1\">transformer.conformer.u2-librispeech\u003C\u002Fa>  \u002F \u003Ca href = \".\u002Fexamples\u002Flibrispeech\u002Fasr2\">transformer.conformer.u2-kaldi-librispeech\u003C\u002Fa>\n      \u003C\u002Ftd>\n      \u003C\u002Ftd>\n    \u003C\u002Ftr>\n  \u003Ctr>\n      \u003Ctd>TIMIT\u003C\u002Ftd>\n      \u003Ctd>Unified Streaming & Non-streaming Two-pass\u003C\u002Ftd>\n      \u003Ctd>\n    \u003Ca href = \".\u002Fexamples\u002Ftimit\u002Fasr1\"> u2-timit\u003C\u002Fa>\n      \u003C\u002Ftd>\n  \u003C\u002Ftr>\n  \u003Ctr>\n  \u003Ctd>Alignment\u003C\u002Ftd>\n  \u003Ctd>THCHS30\u003C\u002Ftd>\n  \u003Ctd>MFA\u003C\u002Ftd>\n  \u003Ctd>\n  \u003Ca href = \".examples\u002Fthchs30\u002Falign0\">mfa-thchs30\u003C\u002Fa>\n  \u003C\u002Ftd>\n  \u003C\u002Ftr>\n   \u003Ctr>\n      \u003Ctd rowspan=\"1\">Language Model\u003C\u002Ftd>\n      \u003Ctd colspan = \"2\">Ngram Language Model\u003C\u002Ftd>\n      \u003Ctd>\n      \u003Ca href = \".\u002Fexamples\u002Fother\u002Fngram_lm\">kenlm\u003C\u002Fa>\n      \u003C\u002Ftd>\n    \u003C\u002Ftr>\n  \u003Ctr>\n      \u003Ctd rowspan=\"2\">Speech Translation (English to Chinese)\u003C\u002Ftd> \n      \u003Ctd rowspan=\"2\">TED En-Zh\u003C\u002Ftd>\n      \u003Ctd>Transformer + ASR MTL\u003C\u002Ftd>\n      \u003Ctd>\n      \u003Ca href = \".\u002Fexamples\u002Fted_en_zh\u002Fst0\">transformer-ted\u003C\u002Fa>\n      \u003C\u002Ftd>\n  \u003C\u002Ftr>\n  \u003Ctr>\n      \u003Ctd>FAT + Transformer + ASR MTL\u003C\u002Ftd>\n      \u003Ctd>\n      \u003Ca href = \".\u002Fexamples\u002Fted_en_zh\u002Fst1\">fat-st-ted\u003C\u002Fa>\n      \u003C\u002Ftd>\n  \u003C\u002Ftr>\n  \u003C\u002Ftbody>\n\u003C\u002Ftable>\n\n\u003Ca name=\"TextToSpeech\">\u003C\u002Fa>\n\n**Text-to-Speech** in PaddleSpeech mainly contains three modules: *Text Frontend*, *Acoustic Model* and *Vocoder*. Acoustic Model and Vocoder models are listed as follow:\n\n\u003Ctable>\n  \u003Cthead>\n    \u003Ctr>\n      \u003Cth> Text-to-Speech Module Type \u003C\u002Fth>\n      \u003Cth> Model Type \u003C\u002Fth>\n      \u003Cth> Dataset \u003C\u002Fth>\n      \u003Cth> Example \u003C\u002Fth>\n    \u003C\u002Ftr>\n  \u003C\u002Fthead>\n  \u003Ctbody>\n    \u003Ctr>\n      \u003Ctd> Text Frontend \u003C\u002Ftd>\n      \u003Ctd colspan=\"2\">   \u003C\u002Ftd>\n      \u003Ctd>\n      \u003Ca href = \".\u002Fexamples\u002Fother\u002Ftn\">tn\u003C\u002Fa> \u002F \u003Ca href = \".\u002Fexamples\u002Fother\u002Fg2p\">g2p\u003C\u002Fa>\n      \u003C\u002Ftd>\n    \u003C\u002Ftr>\n    \u003Ctr>\n      \u003Ctd rowspan=\"6\">Acoustic Model\u003C\u002Ftd>\n      \u003Ctd>Tacotron2\u003C\u002Ftd>\n      \u003Ctd>LJSpeech \u002F CSMSC\u003C\u002Ftd>\n      \u003Ctd>\n      \u003Ca href = \".\u002Fexamples\u002Fljspeech\u002Ftts0\">tacotron2-ljspeech\u003C\u002Fa> \u002F \u003Ca href = \".\u002Fexamples\u002Fcsmsc\u002Ftts0\">tacotron2-csmsc\u003C\u002Fa>\n      \u003C\u002Ftd>\n    \u003C\u002Ftr>\n    \u003Ctr>\n      \u003Ctd>Transformer TTS\u003C\u002Ftd>\n      \u003Ctd>LJSpeech\u003C\u002Ftd>\n      \u003Ctd>\n      \u003Ca href = \".\u002Fexamples\u002Fljspeech\u002Ftts1\">transformer-ljspeech\u003C\u002Fa>\n      \u003C\u002Ftd>\n    \u003C\u002Ftr>\n    \u003Ctr>\n      \u003Ctd>SpeedySpeech\u003C\u002Ftd>\n      \u003Ctd>CSMSC\u003C\u002Ftd>\n      \u003Ctd >\n      \u003Ca href = \".\u002Fexamples\u002Fcsmsc\u002Ftts2\">speedyspeech-csmsc\u003C\u002Fa>\n      \u003C\u002Ftd>\n    \u003C\u002Ftr>\n    \u003Ctr>\n      \u003Ctd>FastSpeech2\u003C\u002Ftd>\n      \u003Ctd>LJSpeech \u002F VCTK \u002F CSMSC \u002F AISHELL-3 \u002F ZH_EN \u002F finetune\u003C\u002Ftd>\n      \u003Ctd>\n      \u003Ca href = \".\u002Fexamples\u002Fljspeech\u002Ftts3\">fastspeech2-ljspeech\u003C\u002Fa> \u002F \u003Ca href = \".\u002Fexamples\u002Fvctk\u002Ftts3\">fastspeech2-vctk\u003C\u002Fa> \u002F \u003Ca href = \".\u002Fexamples\u002Fcsmsc\u002Ftts3\">fastspeech2-csmsc\u003C\u002Fa> \u002F \u003Ca href = \".\u002Fexamples\u002Faishell3\u002Ftts3\">fastspeech2-aishell3\u003C\u002Fa> \u002F \u003Ca href = \".\u002Fexamples\u002Fzh_en_tts\u002Ftts3\">fastspeech2-zh_en\u003C\u002Fa> \u002F \u003Ca href = \".\u002Fexamples\u002Fother\u002Ftts_finetune\u002Ftts3\">fastspeech2-finetune\u003C\u002Fa>\n      \u003C\u002Ftd>\n    \u003C\u002Ftr>\n    \u003Ctr>\n      \u003Ctd>\u003Ca href = \"https:\u002F\u002Farxiv.org\u002Fabs\u002F2211.03545\">ERNIE-SAT\u003C\u002Fa>\u003C\u002Ftd>\n      \u003Ctd>VCTK \u002F AISHELL-3 \u002F ZH_EN\u003C\u002Ftd>\n      \u003Ctd>\n      \u003Ca href = \".\u002Fexamples\u002Fvctk\u002Fernie_sat\">ERNIE-SAT-vctk\u003C\u002Fa> \u002F \u003Ca href = \".\u002Fexamples\u002Faishell3\u002Fernie_sat\">ERNIE-SAT-aishell3\u003C\u002Fa> \u002F \u003Ca href = \".\u002Fexamples\u002Faishell3_vctk\u002Fernie_sat\">ERNIE-SAT-zh_en\u003C\u002Fa>\n      \u003C\u002Ftd>\n    \u003C\u002Ftr>\n    \u003Ctr>\n      \u003Ctd>DiffSinger\u003C\u002Ftd>\n      \u003Ctd>Opencpop\u003C\u002Ftd>\n      \u003Ctd>\n      \u003Ca href = \".\u002Fexamples\u002Fopencpop\u002Fsvs1\">DiffSinger-opencpop\u003C\u002Fa>\n      \u003C\u002Ftd>\n   \u003C\u002Ftr>\n   \u003Ctr>\n      \u003Ctd rowspan=\"6\">Vocoder\u003C\u002Ftd>\n      \u003Ctd >WaveFlow\u003C\u002Ftd>\n      \u003Ctd >LJSpeech\u003C\u002Ftd>\n      \u003Ctd>\n      \u003Ca href = \".\u002Fexamples\u002Fljspeech\u002Fvoc0\">waveflow-ljspeech\u003C\u002Fa>\n      \u003C\u002Ftd>\n    \u003C\u002Ftr>\n    \u003Ctr>\n      \u003Ctd >Parallel WaveGAN\u003C\u002Ftd>\n      \u003Ctd >LJSpeech \u002F VCTK \u002F CSMSC \u002F AISHELL-3 \u002F Opencpop\u003C\u002Ftd>\n      \u003Ctd>\n      \u003Ca href = \".\u002Fexamples\u002Fljspeech\u002Fvoc1\">PWGAN-ljspeech\u003C\u002Fa> \u002F \u003Ca href = \".\u002Fexamples\u002Fvctk\u002Fvoc1\">PWGAN-vctk\u003C\u002Fa> \u002F \u003Ca href = \".\u002Fexamples\u002Fcsmsc\u002Fvoc1\">PWGAN-csmsc\u003C\u002Fa> \u002F  \u003Ca href = \".\u002Fexamples\u002Faishell3\u002Fvoc1\">PWGAN-aishell3\u003C\u002Fa> \u002F \u003Ca href = \".\u002Fexamples\u002Fopencpop\u002Fvoc1\">PWGAN-opencpop\u003C\u002Fa>\n      \u003C\u002Ftd>\n    \u003C\u002Ftr>\n    \u003Ctr>\n      \u003Ctd >Multi Band MelGAN\u003C\u002Ftd>\n      \u003Ctd >CSMSC\u003C\u002Ftd>\n      \u003Ctd>\n      \u003Ca href = \".\u002Fexamples\u002Fcsmsc\u002Fvoc3\">Multi Band MelGAN-csmsc\u003C\u002Fa> \n      \u003C\u002Ftd>\n    \u003C\u002Ftr> \n    \u003Ctr>\n      \u003Ctd >Style MelGAN\u003C\u002Ftd>\n      \u003Ctd >CSMSC\u003C\u002Ftd>\n      \u003Ctd>\n      \u003Ca href = \".\u002Fexamples\u002Fcsmsc\u002Fvoc4\">Style MelGAN-csmsc\u003C\u002Fa> \n      \u003C\u002Ftd>\n    \u003C\u002Ftr>\n    \u003Ctr>\n      \u003Ctd>HiFiGAN\u003C\u002Ftd>\n      \u003Ctd>LJSpeech \u002F VCTK \u002F CSMSC \u002F AISHELL-3 \u002F Opencpop\u003C\u002Ftd>\n      \u003Ctd>\n      \u003Ca href = \".\u002Fexamples\u002Fljspeech\u002Fvoc5\">HiFiGAN-ljspeech\u003C\u002Fa> \u002F \u003Ca href = \".\u002Fexamples\u002Fvctk\u002Fvoc5\">HiFiGAN-vctk\u003C\u002Fa> \u002F \u003Ca href = \".\u002Fexamples\u002Fcsmsc\u002Fvoc5\">HiFiGAN-csmsc\u003C\u002Fa> \u002F \u003Ca href = \".\u002Fexamples\u002Faishell3\u002Fvoc5\">HiFiGAN-aishell3\u003C\u002Fa> \u002F \u003Ca href = \".\u002Fexamples\u002Fopencpop\u002Fvoc5\">HiFiGAN-opencpop\u003C\u002Fa>\n      \u003C\u002Ftd>\n    \u003C\u002Ftr>\n    \u003Ctr>\n      \u003Ctd>WaveRNN\u003C\u002Ftd>\n      \u003Ctd>CSMSC\u003C\u002Ftd>\n      \u003Ctd>\n      \u003Ca href = \".\u002Fexamples\u002Fcsmsc\u002Fvoc6\">WaveRNN-csmsc\u003C\u002Fa>\n      \u003C\u002Ftd>\n    \u003C\u002Ftr>\n    \u003Ctr>\n      \u003Ctd rowspan=\"5\">Voice Cloning\u003C\u002Ftd>\n      \u003Ctd>GE2E\u003C\u002Ftd>\n      \u003Ctd >Librispeech, etc.\u003C\u002Ftd>\n      \u003Ctd>\n      \u003Ca href = \".\u002Fexamples\u002Fother\u002Fge2e\">GE2E\u003C\u002Fa>\n      \u003C\u002Ftd>\n    \u003C\u002Ftr>\n    \u003Ctr>\n      \u003Ctd>SV2TTS (GE2E + Tacotron2)\u003C\u002Ftd>\n      \u003Ctd>AISHELL-3\u003C\u002Ftd>\n      \u003Ctd>\n      \u003Ca href = \".\u002Fexamples\u002Faishell3\u002Fvc0\">VC0\u003C\u002Fa>\n      \u003C\u002Ftd>\n    \u003C\u002Ftr>\n    \u003Ctr>\n      \u003Ctd>SV2TTS (GE2E + FastSpeech2)\u003C\u002Ftd>\n      \u003Ctd>AISHELL-3\u003C\u002Ftd>\n      \u003Ctd>\n      \u003Ca href = \".\u002Fexamples\u002Faishell3\u002Fvc1\">VC1\u003C\u002Fa>\n      \u003C\u002Ftd>\n    \u003C\u002Ftr>\n    \u003Ctr>\n      \u003Ctd>SV2TTS (ECAPA-TDNN + FastSpeech2)\u003C\u002Ftd>\n      \u003Ctd>AISHELL-3\u003C\u002Ftd>\n      \u003Ctd>\n      \u003Ca href = \".\u002Fexamples\u002Faishell3\u002Fvc2\">VC2\u003C\u002Fa>\n      \u003C\u002Ftd>\n    \u003C\u002Ftr>\n    \u003Ctr>\n      \u003Ctd>GE2E + VITS\u003C\u002Ftd>\n      \u003Ctd>AISHELL-3\u003C\u002Ftd>\n      \u003Ctd>\n      \u003Ca href = \".\u002Fexamples\u002Faishell3\u002Fvits-vc\">VITS-VC\u003C\u002Fa>\n      \u003C\u002Ftd>\n    \u003C\u002Ftr>\n    \u003Ctr>\n      \u003Ctd rowspan=\"3\">End-to-End\u003C\u002Ftd>\n      \u003Ctd>VITS\u003C\u002Ftd>\n      \u003Ctd>CSMSC \u002F AISHELL-3\u003C\u002Ftd>\n      \u003Ctd>\n      \u003Ca href = \".\u002Fexamples\u002Fcsmsc\u002Fvits\">VITS-csmsc\u003C\u002Fa> \u002F \u003Ca href = \".\u002Fexamples\u002Faishell3\u002Fvits\">VITS-aishell3\u003C\u002Fa>\n      \u003C\u002Ftd>\n    \u003C\u002Ftr>\n  \u003C\u002Ftbody>\n\u003C\u002Ftable>\n\n\u003Ca name=\"AudioClassification\">\u003C\u002Fa>\n\n**Audio Classification**\n\n\u003Ctable style=\"width:100%\">\n  \u003Cthead>\n    \u003Ctr>\n      \u003Cth> Task \u003C\u002Fth>\n      \u003Cth> Dataset \u003C\u002Fth>\n      \u003Cth> Model Type \u003C\u002Fth>\n      \u003Cth> Example \u003C\u002Fth>\n    \u003C\u002Ftr>\n  \u003C\u002Fthead>\n  \u003Ctbody>\n  \u003Ctr>\n      \u003Ctd>Audio Classification\u003C\u002Ftd>\n      \u003Ctd>ESC-50\u003C\u002Ftd>\n      \u003Ctd>PANN\u003C\u002Ftd>\n      \u003Ctd>\n      \u003Ca href = \".\u002Fexamples\u002Fesc50\u002Fcls0\">pann-esc50\u003C\u002Fa>\n      \u003C\u002Ftd>\n    \u003C\u002Ftr>\n  \u003C\u002Ftbody>\n\u003C\u002Ftable>\n\n\u003Ca name=\"KeywordSpotting\">\u003C\u002Fa>\n\n**Keyword Spotting**\n\n\u003Ctable style=\"width:100%\">\n  \u003Cthead>\n    \u003Ctr>\n      \u003Cth> Task \u003C\u002Fth>\n      \u003Cth> Dataset \u003C\u002Fth>\n      \u003Cth> Model Type \u003C\u002Fth>\n      \u003Cth> Example \u003C\u002Fth>\n    \u003C\u002Ftr>\n  \u003C\u002Fthead>\n  \u003Ctbody>\n  \u003Ctr>\n      \u003Ctd>Keyword Spotting\u003C\u002Ftd>\n      \u003Ctd>hey-snips\u003C\u002Ftd>\n      \u003Ctd>MDTC\u003C\u002Ftd>\n      \u003Ctd>\n      \u003Ca href = \".\u002Fexamples\u002Fhey_snips\u002Fkws0\">mdtc-hey-snips\u003C\u002Fa>\n      \u003C\u002Ftd>\n    \u003C\u002Ftr>\n  \u003C\u002Ftbody>\n\u003C\u002Ftable>\n\n\u003Ca name=\"SpeakerVerification\">\u003C\u002Fa>\n\n**Speaker Verification**\n\n\u003Ctable style=\"width:100%\">\n  \u003Cthead>\n    \u003Ctr>\n      \u003Cth> Task \u003C\u002Fth>\n      \u003Cth> Dataset \u003C\u002Fth>\n      \u003Cth> Model Type \u003C\u002Fth>\n      \u003Cth> Example \u003C\u002Fth>\n    \u003C\u002Ftr>\n  \u003C\u002Fthead>\n  \u003Ctbody>\n  \u003Ctr>\n      \u003Ctd>Speaker Verification\u003C\u002Ftd>\n      \u003Ctd>VoxCeleb1\u002F2\u003C\u002Ftd>\n      \u003Ctd>ECAPA-TDNN\u003C\u002Ftd>\n      \u003Ctd>\n      \u003Ca href = \".\u002Fexamples\u002Fvoxceleb\u002Fsv0\">ecapa-tdnn-voxceleb12\u003C\u002Fa>\n      \u003C\u002Ftd>\n    \u003C\u002Ftr>\n  \u003C\u002Ftbody>\n\u003C\u002Ftable>\n\n\u003Ca name=\"SpeakerDiarization\">\u003C\u002Fa>\n\n**Speaker Diarization**\n\n\u003Ctable style=\"width:100%\">\n  \u003Cthead>\n    \u003Ctr>\n      \u003Cth> Task \u003C\u002Fth>\n      \u003Cth> Dataset \u003C\u002Fth>\n      \u003Cth> Model Type \u003C\u002Fth>\n      \u003Cth> Example \u003C\u002Fth>\n    \u003C\u002Ftr>\n  \u003C\u002Fthead>\n  \u003Ctbody>\n  \u003Ctr>\n      \u003Ctd>Speaker Diarization\u003C\u002Ftd>\n     \u003Ctd>AMI\u003C\u002Ftd>\n      \u003Ctd>ECAPA-TDNN + AHC \u002F SC\u003C\u002Ftd>\n      \u003Ctd>\n      \u003Ca href = \".\u002Fexamples\u002Fami\u002Fsd0\">ecapa-tdnn-ami\u003C\u002Fa>\n      \u003C\u002Ftd>\n    \u003C\u002Ftr>\n  \u003C\u002Ftbody>\n\u003C\u002Ftable>\n\n\u003Ca name=\"PunctuationRestoration\">\u003C\u002Fa>\n\n**Punctuation Restoration**\n\n\u003Ctable style=\"width:100%\">\n  \u003Cthead>\n    \u003Ctr>\n      \u003Cth> Task \u003C\u002Fth>\n      \u003Cth> Dataset \u003C\u002Fth>\n      \u003Cth> Model Type \u003C\u002Fth>\n      \u003Cth> Example \u003C\u002Fth>\n    \u003C\u002Ftr>\n  \u003C\u002Fthead>\n  \u003Ctbody>\n  \u003Ctr>\n      \u003Ctd>Punctuation Restoration\u003C\u002Ftd>\n      \u003Ctd>IWLST2012_zh\u003C\u002Ftd>\n      \u003Ctd>Ernie Linear\u003C\u002Ftd>\n      \u003Ctd>\n      \u003Ca href = \".\u002Fexamples\u002Fiwslt2012\u002Fpunc0\">iwslt2012-punc0\u003C\u002Fa>\n      \u003C\u002Ftd>\n    \u003C\u002Ftr>\n  \u003C\u002Ftbody>\n\u003C\u002Ftable>\n\n## Documents\n\nNormally, [Speech SoTA](https:\u002F\u002Fpaperswithcode.com\u002Farea\u002Fspeech), [Audio SoTA](https:\u002F\u002Fpaperswithcode.com\u002Farea\u002Faudio) and [Music SoTA](https:\u002F\u002Fpaperswithcode.com\u002Farea\u002Fmusic) give you an overview of the hot academic topics in the related area. To focus on the tasks in PaddleSpeech, you will find the following guidelines are helpful to grasp the core ideas.\n\n- [Installation](.\u002Fdocs\u002Fsource\u002Finstall.md)\n- [Quick Start](#quickstart)\n- [Some Demos](.\u002Fdemos\u002FREADME.md)\n- Tutorials\n  - [Automatic Speech Recognition](.\u002Fdocs\u002Fsource\u002Fasr\u002Fquick_start.md)\n    - [Introduction](.\u002Fdocs\u002Fsource\u002Fasr\u002Fmodels_introduction.md)\n    - [Data Preparation](.\u002Fdocs\u002Fsource\u002Fasr\u002Fdata_preparation.md)\n    - [Ngram LM](.\u002Fdocs\u002Fsource\u002Fasr\u002Fngram_lm.md)\n  - [Text-to-Speech](.\u002Fdocs\u002Fsource\u002Ftts\u002Fquick_start.md)\n    - [Introduction](.\u002Fdocs\u002Fsource\u002Ftts\u002Fmodels_introduction.md)\n    - [Advanced Usage](.\u002Fdocs\u002Fsource\u002Ftts\u002Fadvanced_usage.md)\n    - [Chinese Rule Based Text Frontend](.\u002Fdocs\u002Fsource\u002Ftts\u002Fzh_text_frontend.md)\n    - [Test Audio Samples](https:\u002F\u002Fpaddlespeech.readthedocs.io\u002Fen\u002Flatest\u002Ftts\u002Fdemo.html)\n  - Speaker Verification\n    - [Audio Searching](.\u002Fdemos\u002Faudio_searching\u002FREADME.md)\n    - [Speaker Verification](.\u002Fdemos\u002Fspeaker_verification\u002FREADME.md)\n  - [Audio Classification](.\u002Fdemos\u002Faudio_tagging\u002FREADME.md)\n  - [Speech Translation](.\u002Fdemos\u002Fspeech_translation\u002FREADME.md)\n  - [Speech Server](.\u002Fdemos\u002Fspeech_server\u002FREADME.md)\n- [Released Models](.\u002Fdocs\u002Fsource\u002Freleased_model.md)\n  - [Speech-to-Text](#SpeechToText)\n  - [Text-to-Speech](#TextToSpeech)\n  - [Audio Classification](#AudioClassification)\n  - [Speaker Verification](#SpeakerVerification)\n  - [Speaker Diarization](#SpeakerDiarization)\n  - [Punctuation Restoration](#PunctuationRestoration)\n- [Community](#Community)\n- [Welcome to contribute](#contribution)\n- [License](#License)\n\nThe Text-to-Speech module is originally called [Parakeet](https:\u002F\u002Fgithub.com\u002FPaddlePaddle\u002FParakeet), and now merged with this repository. If you are interested in academic research about this task, please see [TTS research overview](https:\u002F\u002Fgithub.com\u002FPaddlePaddle\u002FPaddleSpeech\u002Ftree\u002Fdevelop\u002Fdocs\u002Fsource\u002Ftts#overview). Also, [this document](https:\u002F\u002Fgithub.com\u002FPaddlePaddle\u002FPaddleSpeech\u002Fblob\u002Fdevelop\u002Fdocs\u002Fsource\u002Ftts\u002Fmodels_introduction.md) is a good guideline for the pipeline components.\n\n\n## ⭐ Examples\n- **[PaddleBoBo](https:\u002F\u002Fgithub.com\u002FJiehangXie\u002FPaddleBoBo): Use PaddleSpeech TTS to generate virtual human voice.**\n  \n\u003Cdiv align=\"center\">\u003Ca href=\"https:\u002F\u002Fwww.bilibili.com\u002Fvideo\u002FBV1cL411V71o?share_source=copy_web\">\u003Cimg src=\"https:\u002F\u002Fai-studio-static-online.cdn.bcebos.com\u002F06fd746ab32042f398fb6f33f873e6869e846fe63c214596ae37860fe8103720\" \u002F width=\"500px\">\u003C\u002Fa>\u003C\u002Fdiv>\n\n- [PaddleSpeech Demo Video](https:\u002F\u002Fpaddlespeech.readthedocs.io\u002Fen\u002Flatest\u002Fdemo_video.html)\n\n- **[VTuberTalk](https:\u002F\u002Fgithub.com\u002Fjerryuhoo\u002FVTuberTalk): Use PaddleSpeech TTS and ASR to clone voice from videos.**\n\n\n## Citation\n\nTo cite PaddleSpeech for research, please use the following format.\n\n```text\n@inproceedings{zhang2022paddlespeech,\n    title = {PaddleSpeech: An Easy-to-Use All-in-One Speech Toolkit},\n    author = {Hui Zhang, Tian Yuan, Junkun Chen, Xintong Li, Renjie Zheng, Yuxin Huang, Xiaojie Chen, Enlei Gong, Zeyu Chen, Xiaoguang Hu, dianhai yu, Yanjun Ma, Liang Huang},\n    booktitle = {Proceedings of the 2022 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies: Demonstrations},\n    year = {2022},\n    publisher = {Association for Computational Linguistics},\n}\n\n@InProceedings{pmlr-v162-bai22d,\n  title = {{A}$^3${T}: Alignment-Aware Acoustic and Text Pretraining for Speech Synthesis and Editing},\n  author = {Bai, He and Zheng, Renjie and Chen, Junkun and Ma, Mingbo and Li, Xintong and Huang, Liang},\n  booktitle = {Proceedings of the 39th International Conference on Machine Learning},\n  pages = {1399--1411},\n  year = {2022},\n  volume = {162},\n  series = {Proceedings of Machine Learning Research},\n  month = {17--23 Jul},\n  publisher = {PMLR},\n  pdf = {https:\u002F\u002Fproceedings.mlr.press\u002Fv162\u002Fbai22d\u002Fbai22d.pdf},\n  url = {https:\u002F\u002Fproceedings.mlr.press\u002Fv162\u002Fbai22d.html},\n}\n\n@inproceedings{zheng2021fused,\n  title={Fused acoustic and text encoding for multimodal bilingual pretraining and speech translation},\n  author={Zheng, Renjie and Chen, Junkun and Ma, Mingbo and Huang, Liang},\n  booktitle={International Conference on Machine Learning},\n  pages={12736--12746},\n  year={2021},\n  organization={PMLR}\n}\n```\n\n\u003Ca name=\"contribution\">\u003C\u002Fa>\n## Contribute to PaddleSpeech\n\nYou are warmly welcome to submit questions in [discussions](https:\u002F\u002Fgithub.com\u002FPaddlePaddle\u002FPaddleSpeech\u002Fdiscussions) and bug reports in [issues](https:\u002F\u002Fgithub.com\u002FPaddlePaddle\u002FPaddleSpeech\u002Fissues)! Also, we highly appreciate if you are willing to contribute to this project!\n\n### Contributors\n\u003Cp align=\"center\">\n\u003Ca href=\"https:\u002F\u002Fgithub.com\u002Fzh794390558\">\u003Cimg src=\"https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F3038472?s=60&v=4\" width=75 height=75>\u003C\u002Fa>\n\u003Ca href=\"https:\u002F\u002Fgithub.com\u002FJackwaterveg\">\u003Cimg src=\"https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F87408988?s=60&v=4\" width=75 height=75>\u003C\u002Fa>\n\u003Ca href=\"https:\u002F\u002Fgithub.com\u002Fyt605155624\">\u003Cimg src=\"https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F24568452?s=60&v=4\" width=75 height=75>\u003C\u002Fa>\n\u003Ca href=\"https:\u002F\u002Fgithub.com\u002FHonei\">\u003Cimg src=\"https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F11361692?s=60&v=4\" width=75 height=75>\u003C\u002Fa>\n\u003Ca href=\"https:\u002F\u002Fgithub.com\u002FKPatr1ck\">\u003Cimg src=\"https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F22954146?s=60&v=4\" width=75 height=75>\u003C\u002Fa>\n\u003Ca href=\"https:\u002F\u002Fgithub.com\u002Fkuke\">\u003Cimg src=\"https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F3064195?s=60&v=4\" width=75 height=75>\u003C\u002Fa>\n\u003Ca href=\"https:\u002F\u002Fgithub.com\u002Flym0302\">\u003Cimg src=\"https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F34430015?s=60&v=4\" width=75 height=75>\u003C\u002Fa>\n\u003Ca href=\"https:\u002F\u002Fgithub.com\u002FSmileGoat\">\u003Cimg src=\"https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F56786796?s=60&v=4\" width=75 height=75>\u003C\u002Fa>\n\u003Ca href=\"https:\u002F\u002Fgithub.com\u002Fxinghai-sun\">\u003Cimg src=\"https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F7038341?s=60&v=4\" width=75 height=75>\u003C\u002Fa>\n\u003Ca href=\"https:\u002F\u002Fgithub.com\u002Fpkuyym\">\u003Cimg src=\"https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F5782283?s=60&v=4\" width=75 height=75>\u003C\u002Fa>\n\u003Ca href=\"https:\u002F\u002Fgithub.com\u002FLittleChenCc\">\u003Cimg src=\"https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F10339970?s=60&v=4\" width=75 height=75>\u003C\u002Fa>\n\u003Ca href=\"https:\u002F\u002Fgithub.com\u002Fqingen\">\u003Cimg src=\"https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F3139179?s=60&v=4\" width=75 height=75>\u003C\u002Fa>\n\u003Ca href=\"https:\u002F\u002Fgithub.com\u002FD-DanielYang\">\u003Cimg src=\"https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F23690325?s=60&v=4\" width=75 height=75>\u003C\u002Fa>\n\u003Ca href=\"https:\u002F\u002Fgithub.com\u002FMingxue-Xu\">\u003Cimg src=\"https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F92848346?s=60&v=4\" width=75 height=75>\u003C\u002Fa>\n\u003Ca href=\"https:\u002F\u002Fgithub.com\u002F745165806\">\u003Cimg src=\"https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F20623194?s=60&v=4\" width=75 height=75>\u003C\u002Fa>\n\u003Ca href=\"https:\u002F\u002Fgithub.com\u002Fjerryuhoo\">\u003Cimg src=\"https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F24245709?s=60&v=4\" width=75 height=75>\u003C\u002Fa>\n\u003Ca href=\"https:\u002F\u002Fgithub.com\u002FWilliamZhang06\">\u003Cimg src=\"https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F97937340?s=60&v=4\" width=75 height=75>\u003C\u002Fa>\n\u003Ca href=\"https:\u002F\u002Fgithub.com\u002Fchrisxu2016\">\u003Cimg src=\"https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F18379485?s=60&v=4\" width=75 height=75>\u003C\u002Fa>\n\u003Ca href=\"https:\u002F\u002Fgithub.com\u002Fiftaken\">\u003Cimg src=\"https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F30135920?s=60&v=4\" width=75 height=75>\u003C\u002Fa>\n\u003Ca href=\"https:\u002F\u002Fgithub.com\u002Flfchener\">\u003Cimg src=\"https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F6771821?s=60&v=4\" width=75 height=75>\u003C\u002Fa>\n\u003Ca href=\"https:\u002F\u002Fgithub.com\u002FBarryKCL\">\u003Cimg src=\"https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F48039828?s=60&v=4\" width=75 height=75>\u003C\u002Fa>\n\u003Ca href=\"https:\u002F\u002Fgithub.com\u002Fmmglove\">\u003Cimg src=\"https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F38800877?s=60&v=4\" width=75 height=75>\u003C\u002Fa>\n\u003Ca href=\"https:\u002F\u002Fgithub.com\u002Fgongel\">\u003Cimg src=\"https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F24390500?s=60&v=4\" width=75 height=75>\u003C\u002Fa>\n\u003Ca href=\"https:\u002F\u002Fgithub.com\u002Fluotao1\">\u003Cimg src=\"https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F6836917?s=60&v=4\" width=75 height=75>\u003C\u002Fa>\n\u003Ca href=\"https:\u002F\u002Fgithub.com\u002Fwanghaoshuang\">\u003Cimg src=\"https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F7534971?s=60&v=4\" width=75 height=75>\u003C\u002Fa>\n\u003Ca href=\"https:\u002F\u002Fgithub.com\u002Fkslz\">\u003Cimg src=\"https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F54951765?s=60&v=4\" width=75 height=75>\u003C\u002Fa>\n\u003Ca href=\"https:\u002F\u002Fgithub.com\u002FJiehangXie\">\u003Cimg src=\"https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F51190264?s=60&v=4\" width=75 height=75>\u003C\u002Fa>\n\u003Ca href=\"https:\u002F\u002Fgithub.com\u002Fdavid-95\">\u003Cimg src=\"https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F15189190?s=60&v=4\" width=75 height=75>\u003C\u002Fa>\n\u003Ca href=\"https:\u002F\u002Fgithub.com\u002FTHUzyt21\">\u003Cimg src=\"https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F91456992?s=60&v=4\" width=75 height=75>\u003C\u002Fa>\n\u003Ca href=\"https:\u002F\u002Fgithub.com\u002Fbuchongyu2\">\u003Cimg src=\"https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F29157444?s=60&v=4\" width=75 height=75>\u003C\u002Fa>\n\u003Ca href=\"https:\u002F\u002Fgithub.com\u002Ficlementine\">\u003Cimg src=\"https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F16222986?s=60&v=4\" width=75 height=75>\u003C\u002Fa>\n\u003Ca href=\"https:\u002F\u002Fgithub.com\u002Fphecda-xu\">\u003Cimg src=\"https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F46859427?s=60&v=4\" width=75 height=75>\u003C\u002Fa>\n\u003Ca href=\"https:\u002F\u002Fgithub.com\u002Ffreeliuzc\">\u003Cimg src=\"https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F23568094?s=60&v=4\" width=75 height=75>\u003C\u002Fa>\n\u003Ca href=\"https:\u002F\u002Fgithub.com\u002FZeyuChen\">\u003Cimg src=\"https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F1371212?s=60&v=4\" width=75 height=75>\u003C\u002Fa>\n\u003Ca href=\"https:\u002F\u002Fgithub.com\u002Fccrrong\">\u003Cimg src=\"https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F101700995?s=60&v=4\" width=75 height=75>\u003C\u002Fa>\n\u003Ca href=\"https:\u002F\u002Fgithub.com\u002FAK391\">\u003Cimg src=\"https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F81195143?s=60&v=4\" width=75 height=75>\u003C\u002Fa>\n\u003Ca href=\"https:\u002F\u002Fgithub.com\u002Fqingqing01\">\u003Cimg src=\"https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F7845005?s=60&v=4\" width=75 height=75>\u003C\u002Fa>\n\u003Ca href=\"https:\u002F\u002Fgithub.com\u002F0x45f\">\u003Cimg src=\"https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F23097963?s=60&v=4\" width=75 height=75>\u003C\u002Fa>\n\u003Ca href=\"https:\u002F\u002Fgithub.com\u002Fvpegasus\">\u003Cimg src=\"https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F22723154?s=60&v=4\" width=75 height=75>\u003C\u002Fa>\n\u003Ca href=\"https:\u002F\u002Fgithub.com\u002Fericxk\">\u003Cimg src=\"https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F4719594?s=60&v=4\" width=75 height=75>\u003C\u002Fa>\n\u003Ca href=\"https:\u002F\u002Fgithub.com\u002FBetterman-qs\">\u003Cimg src=\"https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F61459181?s=60&v=4\" width=75 height=75>\u003C\u002Fa>\n\u003Ca href=\"https:\u002F\u002Fgithub.com\u002Fsneaxiy\">\u003Cimg src=\"https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F32832641?s=60&v=4\" width=75 height=75>\u003C\u002Fa>\n\u003Ca href=\"https:\u002F\u002Fgithub.com\u002FDoubledongli\">\u003Cimg src=\"https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F20540661?s=60&v=4\" width=75 height=75>\u003C\u002Fa>\n\u003Ca href=\"https:\u002F\u002Fgithub.com\u002Fapps\u002Fdependabot\">\u003Cimg src=\"https:\u002F\u002Favatars.githubusercontent.com\u002Fin\u002F29110?s=60&v=4\" width=75 height=75>\u003C\u002Fa>\n\u003Ca href=\"https:\u002F\u002Fgithub.com\u002Fkvinwang\">\u003Cimg src=\"https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F6442159?s=60&v=4\" width=75 height=75>\u003C\u002Fa>\n\u003Ca href=\"https:\u002F\u002Fgithub.com\u002Fchenkui164\">\u003Cimg src=\"https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F34813030?s=60&v=4\" width=75 height=75>\u003C\u002Fa>\n\u003Ca href=\"https:\u002F\u002Fgithub.com\u002FPaddleZhang\">\u003Cimg src=\"https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F97284124?s=60&v=4\" width=75 height=75>\u003C\u002Fa>\n\u003Ca href=\"https:\u002F\u002Fgithub.com\u002Fbillishyahao\">\u003Cimg src=\"https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F96406262?s=60&v=4\" width=75 height=75>\u003C\u002Fa>\n\u003Ca href=\"https:\u002F\u002Fgithub.com\u002FBrightXiaoHan\">\u003Cimg src=\"https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F25839309?s=60&v=4\" width=75 height=75>\u003C\u002Fa>\n\u003Ca href=\"https:\u002F\u002Fgithub.com\u002Fjiqiren11\">\u003Cimg src=\"https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F82639260?s=60&v=4\" width=75 height=75>\u003C\u002Fa>\n\u003Ca href=\"https:\u002F\u002Fgithub.com\u002Fryanrussell\">\u003Cimg src=\"https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F523300?s=60&v=4\" width=75 height=75>\u003C\u002Fa>\n\u003Ca href=\"https:\u002F\u002Fgithub.com\u002FGT-ZhangAcer\">\u003Cimg src=\"https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F46156734?s=60&v=4\" width=75 height=75>\u003C\u002Fa>\n\u003Ca href=\"https:\u002F\u002Fgithub.com\u002Ftensor-tang\">\u003Cimg src=\"https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F21351065?s=60&v=4\" width=75 height=75>\u003C\u002Fa>\n\u003Ca href=\"https:\u002F\u002Fgithub.com\u002Fhysunflower\">\u003Cimg src=\"https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F52739577?s=60&v=4\" width=75 height=75>\u003C\u002Fa>\n\u003Ca href=\"https:\u002F\u002Fgithub.com\u002Foyjxer\">\u003Cimg src=\"https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F16233945?s=60&v=4\" width=75 height=75>\u003C\u002Fa>\n\u003Ca href=\"https:\u002F\u002Fgithub.com\u002FJamesLim-sy\">\u003Cimg src=\"https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F61349199?s=60&v=4\" width=75 height=75>\u003C\u002Fa>\n\u003Ca href=\"https:\u002F\u002Fgithub.com\u002Flimpidezza\">\u003Cimg src=\"https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F71760778?s=60&v=4\" width=75 height=75>\u003C\u002Fa>\n\u003Ca href=\"https:\u002F\u002Fgithub.com\u002Fwindstamp\">\u003Cimg src=\"https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F34057289?s=60&v=4\" width=75 height=75>\u003C\u002Fa>\n\u003Ca href=\"https:\u002F\u002Fgithub.com\u002FAshishKarel\">\u003Cimg src=\"https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F58069375?s=60&v=4\" width=75 height=75>\u003C\u002Fa>\n\u003Ca href=\"https:\u002F\u002Fgithub.com\u002Fchesterkuo\">\u003Cimg src=\"https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F6285069?s=60&v=4\" width=75 height=75>\u003C\u002Fa>\n\u003Ca href=\"https:\u002F\u002Fgithub.com\u002FYDX-2147483647\">\u003Cimg src=\"https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F73375426?s=60&v=4\" width=75 height=75>\u003C\u002Fa>\n\u003Ca href=\"https:\u002F\u002Fgithub.com\u002FAdamBear\">\u003Cimg src=\"https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F2288870?s=60&v=4\" width=75 height=75>\u003C\u002Fa>\n\u003Ca href=\"https:\u002F\u002Fgithub.com\u002Fwwhu\">\u003Cimg src=\"https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F6081200?s=60&v=4\" width=75 height=75>\u003C\u002Fa>\n\u003Ca href=\"https:\u002F\u002Fgithub.com\u002Flispc\">\u003Cimg src=\"https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F2833376?s=60&v=4\" width=75 height=75>\u003C\u002Fa>\n\u003Ca href=\"https:\u002F\u002Fgithub.com\u002Fharisankarh\">\u003Cimg src=\"https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F1307053?s=60&v=4\" width=75 height=75>\u003C\u002Fa>\n\u003Ca href=\"https:\u002F\u002Fgithub.com\u002Fpengzhendong\">\u003Cimg src=\"https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F10704539?s=60&v=4\" width=75 height=75>\u003C\u002Fa>\n\u003Ca href=\"https:\u002F\u002Fgithub.com\u002FJackiexiao\">\u003Cimg src=\"https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F18050469?s=60&v=4\" width=75 height=75>\u003C\u002Fa>\n\u003C\u002Fp>\n\n## Acknowledgement\n- Many thanks to [HighCWu](https:\u002F\u002Fgithub.com\u002FHighCWu) for adding [VITS-aishell3](.\u002Fexamples\u002Faishell3\u002Fvits) and [VITS-VC](.\u002Fexamples\u002Faishell3\u002Fvits-vc) examples.\n- Many thanks to [david-95](https:\u002F\u002Fgithub.com\u002Fdavid-95) for fixing multi-punctuation bug、contributing to multiple program and data, and adding [SSML](https:\u002F\u002Fgithub.com\u002FPaddlePaddle\u002FPaddleSpeech\u002Fdiscussions\u002F2538) for TTS Chinese Text Frontend. \n- Many thanks to [BarryKCL](https:\u002F\u002Fgithub.com\u002FBarryKCL) for improving TTS Chinses Frontend based on [G2PW](https:\u002F\u002Fgithub.com\u002FGitYCC\u002Fg2pW).\n- Many thanks to [yeyupiaoling](https:\u002F\u002Fgithub.com\u002Fyeyupiaoling)\u002F[PPASR](https:\u002F\u002Fgithub.com\u002Fyeyupiaoling\u002FPPASR)\u002F[PaddlePaddle-DeepSpeech](https:\u002F\u002Fgithub.com\u002Fyeyupiaoling\u002FPaddlePaddle-DeepSpeech)\u002F[VoiceprintRecognition-PaddlePaddle](https:\u002F\u002Fgithub.com\u002Fyeyupiaoling\u002FVoiceprintRecognition-PaddlePaddle)\u002F[AudioClassification-PaddlePaddle](https:\u002F\u002Fgithub.com\u002Fyeyupiaoling\u002FAudioClassification-PaddlePaddle) for years of attention, constructive advice and great help.\n- Many thanks to [mymagicpower](https:\u002F\u002Fgithub.com\u002Fmymagicpower) for the Java implementation of ASR upon [short](https:\u002F\u002Fgithub.com\u002Fmymagicpower\u002FAIAS\u002Ftree\u002Fmain\u002F3_audio_sdks\u002Fasr_sdk) and [long](https:\u002F\u002Fgithub.com\u002Fmymagicpower\u002FAIAS\u002Ftree\u002Fmain\u002F3_audio_sdks\u002Fasr_long_audio_sdk) audio files.\n- Many thanks to [JiehangXie](https:\u002F\u002Fgithub.com\u002FJiehangXie)\u002F[PaddleBoBo](https:\u002F\u002Fgithub.com\u002FJiehangXie\u002FPaddleBoBo) for developing Virtual Uploader(VUP)\u002FVirtual YouTuber(VTuber) with PaddleSpeech TTS function.\n- Many thanks to [745165806](https:\u002F\u002Fgithub.com\u002F745165806)\u002F[PaddleSpeechTask](https:\u002F\u002Fgithub.com\u002F745165806\u002FPaddleSpeechTask) for contributing Punctuation Restoration model.\n- Many thanks to [kslz](https:\u002F\u002Fgithub.com\u002F745165806) for supplementary Chinese documents.\n- Many thanks to [awmmmm](https:\u002F\u002Fgithub.com\u002Fawmmmm) for contributing fastspeech2 aishell3 conformer pretrained model.\n- Many thanks to [phecda-xu](https:\u002F\u002Fgithub.com\u002Fphecda-xu)\u002F[PaddleDubbing](https:\u002F\u002Fgithub.com\u002Fphecda-xu\u002FPaddleDubbing) for developing a dubbing tool with GUI based on PaddleSpeech TTS model.\n- Many thanks to [jerryuhoo](https:\u002F\u002Fgithub.com\u002Fjerryuhoo)\u002F[VTuberTalk](https:\u002F\u002Fgithub.com\u002Fjerryuhoo\u002FVTuberTalk) for developing a GUI tool based on PaddleSpeech TTS and code for making datasets from videos based on PaddleSpeech ASR.\n- Many thanks to [vpegasus](https:\u002F\u002Fgithub.com\u002Fvpegasus)\u002F[xuesebot](https:\u002F\u002Fgithub.com\u002Fvpegasus\u002Fxuesebot) for developing a rasa chatbot,which is able to speak and listen thanks to PaddleSpeech.\n- Many thanks to [chenkui164](https:\u002F\u002Fgithub.com\u002Fchenkui164)\u002F[FastASR](https:\u002F\u002Fgithub.com\u002Fchenkui164\u002FFastASR) for the C++ inference implementation of PaddleSpeech ASR.\n- Many thanks to [heyudage](https:\u002F\u002Fgithub.com\u002Fheyudage)\u002F[VoiceTyping](https:\u002F\u002Fgithub.com\u002Fheyudage\u002FVoiceTyping) for the real-time voice typing tool implementation of PaddleSpeech ASR streaming services.\n- Many thanks to [EscaticZheng](https:\u002F\u002Fgithub.com\u002FEscaticZheng)\u002F[ps3.9wheel-install](https:\u002F\u002Fgithub.com\u002FEscaticZheng\u002Fps3.9wheel-install) for the python3.9 prebuilt wheel for PaddleSpeech installation in Windows without Visual Studio.\nBesides, PaddleSpeech depends on a lot of open source repositories. See [references](.\u002Fdocs\u002Fsource\u002Freference.md) for more information.\n- Many thanks to [chinobing](https:\u002F\u002Fgithub.com\u002Fchinobing)\u002F[FastAPI-PaddleSpeech-Audio-To-Text](https:\u002F\u002Fgithub.com\u002Fchinobing\u002FFastAPI-PaddleSpeech-Audio-To-Text) for converting audio to text based on FastAPI and PaddleSpeech.\n- Many thanks to [MistEO](https:\u002F\u002Fgithub.com\u002FMistEO)\u002F[Pallas-Bot](https:\u002F\u002Fgithub.com\u002FMistEO\u002FPallas-Bot) for QQ bot based on PaddleSpeech TTS.\n\n\u003Ca name=\"License\">\u003C\u002Fa>\n## License\n\nPaddleSpeech is provided under the [Apache-2.0 License](.\u002FLICENSE).\n\n## Stargazers over time\n\n[![Stargazers over time](https:\u002F\u002Fstarchart.cc\u002FPaddlePaddle\u002FPaddleSpeech.svg)](https:\u002F\u002Fstarchart.cc\u002FPaddlePaddle\u002FPaddleSpeech)\n","PaddlePaddle\u002FPaddleSpeech 是一个基于 PaddlePaddle 平台的开源语音工具包，涵盖了多种关键的语音和音频任务。它提供了包括自监督学习模型、带标点的流式自动语音识别（ASR）、带文本前端的流式文本转语音（TTS）、说话人验证系统、端到端语音翻译和关键词检测等在内的功能。项目采用了诸如 Conformer 和 Transformer 等先进的技术，并支持 Wav2Vec2 和 Whisper 模型。PaddleSpeech 适用于需要高质量语音处理的应用场景，例如智能助手、语音识别应用、语音合成服务以及多语言翻译系统。",2,"2026-06-11 03:34:04","high_star"]