Enhanced English speech corpus for TTS (585 hours)
About 585 hours of clean multi-speaker English speech (24 kHz) with texts for speech synthesis.
- Text → speech
- English
Hugging Face datasets with plain-language descriptions: what's inside and what models to train on them. 4,788 in the catalog so far, with new ones added every hour.
Answered by Gemini 4 · the chat is kept only in this tab
About 585 hours of clean multi-speaker English speech (24 kHz) with texts for speech synthesis.
Part of the large LAION-Audio-630K corpus with English audio-text pairs.
Large audio dataset with transcripts in 22 Indian languages for training text-to-speech systems.
1,000 English audio recordings with complex logical reasoning tasks for evaluating multimodal models.
Multilingual dataset of live dialogues of up to 4 people with separate audio channels for ASR.
Audio questions across 26 locales to evaluate ASR, audio search, and audio embeddings.
A large audio corpus with text-reading recordings and annotations for training ASR and speech classification systems.
Audio recordings of utterances with banking assistant intents in multiple languages.
Audiobooks, podcasts, and YouTube with transcripts for training English ASR and TTS.
Audio and text for expressive speech synthesis in six Indian languages.
403,437 wav voice lines of Honkai: Star Rail characters in Chinese, English, Japanese, and Korean.
Audio dataset of French speech captured by body conduction sensors for recognition, synthesis, and audio enhancement tasks.
30k hours of auto-transcribed speech and a cleaned 22k-hour set for ASR in three languages.
Recordings by famous storyteller Zhang Yuekai with verified texts for Cantonese TTS and ASR.
First multi-dialect dataset of synthetic and real Arabic speech for detecting audio fakes.
Nearly 728 hours of Khmer audio recordings with manual transcription for speech recognition.
Spontaneous speech recordings and transcriptions for 348 languages for ASR training and language identification.
Japanese ASMR audio with images for speech recognition and synthesis; updates are paused.
Classic audio recording and text label dataset for environmental sound recognition.
1,684 hours of manually transcribed speech audio across 12 Indian languages for ASR training.
1000 hours of dual-channel English conversations at 48 kHz for full-duplex and speech models.
English audio and text data for evaluating voice language models.
Up to 950K hours of open speech across 24 official EU languages with automatic transcripts.
A parallel corpus of Arabic Quran recitation audio recordings and corresponding texts for speech recognition tasks.