Speech Recognition Datasets
There are 12 speech recognition datasets in our directory, 1 of which are benchmarks. Each links to its source, paper, and download — browse the full list below or filter by language.
Speech Recognition is the task of transcribing spoken audio into written text. We catalog 12 datasets for it.
- Argentinian Spanish [es-ar] Speech Multi-Speaker DatasetSpeech RecognitionSpanish (Argentinan)
- AudioSetSpeech Recognition, VisualMulti-Lingual
- CMU_ARCTICSpeech RecognitionEnglishBenchmark
- Common VoiceSpeech RecognitionMulti-Lingual
- L2-ARTICSpeech RecognitionEnglish (Non-Native)
- Libri-LightSpeech RecognitionEnglish
- LibriSpeech ASRSpeech RecognitionEnglish
- Microsoft Information-Seeking Conversation (MISC) datasetSpeech Recognition, Dialogue, VisualEnglish
- Microsoft Speech CorpusSpeech RecognitionIndian
- Microsoft Speech Language Translation Corpus (MSLT)Speech Recognition, Machine TranslationMulti-Lingual
- Voices Obscured in Complex Environmental Settings (VOiCES)Speech RecognitionEnglish
- VoxCelebSpeech Recognition, VisualMulti-Lingual