Speech & Audio datasets
46 public Speech & Audio datasets in the Scigantic catalog, from providers including Meta AI, NVIDIA and ESPnet / CMU WAVLab. 46 can be read without an account, and 42 carry an open license (3 are non-commercial or no-derivatives). Common formats: Parquet, CSV and JSON. Each opens in a hosted Jupyter notebook with the data mounted, so nothing has to be downloaded first.
| Dataset | Provider | License | Access |
|---|---|---|---|
| 2 M Belebele Parquet | Meta AI | cc-by-sa-4.0 | public |
| 2 M Flores ASL CSV | Meta AI | cc-by-sa-4.0 | public |
| AMI Meeting Corpus (Kaldi-segmented ASR release) Parquet | AMI Consortium; University of Edinburgh CSTR | CC-BY-4.0 | public |
| ASCEND: spontaneous Chinese-English code-switching speech corpus Parquet | Holy Lovenia, Samuel Cahyawijaya, Genta Indra Winata et al. | CC-BY-SA-4.0 | public |
| Bach Doodle Dataset: 21.6 Million User Melodies with Coconet Four-Voice Bach-Style Harmonizations (Google Magenta) | Google Magenta (Cheng-Zhi Anna Huang et al.) | CC-BY-4.0 | public |
| CML-TTS multilingual speech synthesis dataset Parquet | Frederico S. Oliveira et al., Center of Excellence in Artifi | CC-BY-4.0 | public |
| CocoChorales: 240,000 Synthesized Four-Part Chorale Mixtures with Per-Instrument Stems, MIDI and Note Expression (Google Magenta) | Google Magenta (Yusong Wu, Josh Gardner, Ethan Manilow, Ian | CC-BY-4.0 | public |
| Crowdsourced UK and Ireland English Dialect Speech (OpenSLR 83) Parquet | Isin Demirsahin, Oddur Kjartansson, Alexander Gutkin, Clara | CC-BY-SA-4.0 | public |
| CVSS: Multilingual Speech-to-Speech Translation Corpus, 21 Languages into English (CVSS-C and CVSS-T, Google Research) | Google Research (Ye Jia, Michelle Tadmor Ramanovich, Quan Wa | CC-BY-4.0 | public |
| E-GMD: Expanded Groove MIDI Dataset, 444 Hours of Human Drumming on 43 Drum Kits with Aligned MIDI and Audio (Google Magenta) CSV | Google Magenta (Lee Callender, Curtis Hawthorne, Jesse Engel | CC-BY-4.0 | public |
| Earnings-21: entity-dense earnings call ASR benchmark JSON Lines | Miguel Del Rio, Natalie Delworth, Ryan Westerman et al. (Rev | CC-BY-SA-4.0 | public |
| Earnings-22: accented earnings call ASR benchmark Parquet | Miguel Del Rio, Peter Ha, Quinten McNamara, Corey Miller, Sh | CC-BY-SA-4.0 | public |
| EdAcc: Edinburgh International Accents of English Corpus Parquet | Ramon Sanabria, Nikolay Bogoychev, Nina Markl, Andrea Carman | CC-BY-SA-4.0 | public |
| FLEURS (Few-shot Learning Evaluation of Universal Representations of Speech) TSVParquet | cc-by-4.0 | public | |
| FLORAS: long-form speech recognition, translation and summarization benchmark Parquet | ESPnet / CMU WAVLab | CC-BY-3.0 | public |
| Granary JSON Lines | NVIDIA | cc-by-3.0 | public |
| Groove MIDI Dataset (GMD): 13.6 Hours of Expressive Human Drumming as MIDI with Synthesized Audio (Google Magenta) | Google Magenta (Jon Gillick, Adam Roberts, Jesse Engel, Doug | CC-BY-4.0 | public |
| Hi-Fi Multi-Speaker English TTS Dataset (Hi-Fi TTS) Parquet | Evelina Bakhturina, Vitaly Lavrukhin, Boris Ginsburg, Yang Z | CC-BY-4.0 | public |
| Hifitts 2 JSON | NVIDIA | cc-by-4.0 | public |
| LibriSpeech ASR corpus Parquet | Vassil Panayotov, Guoguo Chen, Daniel Povey, Sanjeev Khudanp | CC-BY-4.0 | public |
| LibriSpeech-Long: Long-Form Speech Benchmark of Contiguous Audiobook Segments up to 4 Minutes (Google DeepMind) | Google DeepMind (Se Jin Park, Julian Salazar, Aren Jansen et | CC-BY-4.0 | public |
| LibriTTS multi-speaker English text-to-speech corpus Parquet | Heiga Zen, Viet Dang, Rob Clark, Yu Zhang, Ron J. Weiss, Ye | CC-BY-4.0 | public |
| LibriTTS-R: restored LibriTTS speech corpus Parquet | Yuma Koizumi, Heiga Zen, Shigeki Karita et al. (Google) | CC-BY-4.0 | public |
| LoquaciousSet (LargeScaleASR): 25,000 hours of English speech Parquet | Titouan Parcollet, Yuan Tseng, Shucong Zhang, Rogier van Dal | CC-BY-4.0 | public |
| MInDS-14: multilingual spoken intent detection Parquet | Daniela Gerz, Pei-Hao Su, Razvan Kusztos, Avishek Mondal, Mi | CC-BY-4.0 | public |
| MMOU JSON | NVIDIA | apache-2.0 | public |
| Multilingual Librispeech Parquet | Meta AI | cc-by-4.0 | public |
| Multilingual Spoken Words Corpus (MSWC) | Mark Mazumder, Sharad Chitlangia, Colby Banbury et al. (MLCo | CC-BY-4.0 | public |
| Nemotron Content Safety Audio Dataset WAVCSV | NVIDIA | cc-by-4.0 | public |
| Norwegian Parliamentary Speech Corpus (NPSC) JSON | Norwegian Language Bank (Sprakbanken), National Library of N | CC-BY-SA-3.0 | public |
| NSynth: Neural Audio Synthesis musical notes dataset Parquet | Jesse Engel, Cinjon Resnick, Adam Roberts, Sander Dieleman, | CC-BY-4.0 | public |
| Omnilingual Asr Corpus Parquet | Meta AI | cc-by-4.0 | public |
| RixVox-v2: Swedish parliamentary speech dataset Parquet | KBLab, National Library of Sweden | ODC-By-1.0 | public |
| speechocean762: non-native English pronunciation scoring corpus Parquet | Junbo Zhang et al. (Xiaomi, SpeechOcean) | CC-BY-4.0 | public |
| The People's Speech: 30,000-hour English speech recognition corpus Parquet | MLCommons (Daniel Galvez et al.) | CC-BY-SA-4.0 | public |
| Upper Waiakea Forest Reserve Passive Acoustic Recordings of Hawaiian Forest Birds WAVCSV | Imageomics Institute and ABC Global Center (Banerji N., Prov | CC-BY-4.0 | public |
| Vibravox: French speech recorded with body-conduction sensors Parquet | Julien Hauret, Malo Olivier, Thomas Joubaud, Christophe Lang | CC-BY-4.0 | public |
| VoiceBank+DEMAND noisy speech database (16 kHz) Parquet | Cassia Valentini-Botinhao (University of Edinburgh CSTR); 16 | CC-BY-4.0 | public |
| Voxpopuli Parquet | Meta AI | cc0-1.0 | public |
| WAXAL: multilingual African language speech corpus (ASR and TTS) JSONParquet | Google, with Makerere University, University of Ghana, Digit | CC-BY-SA-4.0 | public |
| YODAS v3: 1.1-million-hour multilingual web speech corpus Parquet | ESPnet / CMU WAVLab | CC-BY-3.0 | public |
| YODAS2: YouTube-Oriented Dataset for Audio and Speech (long-form) JSON | Xinjian Li, Shinnosuke Takamichi, Takaaki Saeki et al. (CMU | CC-BY-3.0 | public |
| Covost 2 | Meta AI | cc-by-nc-4.0 non-commercial or no-derivatives | public |
| LJ Speech single-speaker English TTS dataset Parquet | Keith Ito, Linda Johnson | Public Domain | public |
| Numb 3 Rs WAVJSON Lines | NVIDIA | cc-by-nc-sa-4.0 non-commercial or no-derivatives | public |
| Seamless Interaction CSV | Meta AI | cc-by-nc-4.0 non-commercial or no-derivatives | public |
Related topics
- Parquet datasets (903)
- Speech recognition datasets (25)
- Benchmark datasets (371)
- Text-to-speech datasets (11)
- JSON datasets (679)
- CSV datasets (651)
- JSON Lines datasets (116)
- WAV datasets (24)