Speech Dataset: Multi-Task Audio for ASR and Synthesis

This collection of speech processing data, updated in January 2026 by double22a, focuses on automatic speech recognition and synthesis tasks. It includes audio files in .wav format and supports diverse applications such as speech-diarization and voice conversion.

Use Cases

• Training automatic speech recognition (ASR) models using the provided .wav audio files
• Developing text-to-speech (TTS) synthesis pipelines to generate artificial voices
• Evaluating speech-separation algorithms to isolate individual speakers from mixed audio

Strengths

• Licensed under Apache-2.0 for open-source and commercial use
• Includes .wav format files for high-fidelity audio processing
• Covers specialized speech tasks including diarization and enhancement

Limitations

• Unknown record count and total audio duration
• Lack of documented column metadata or schema
• No information provided regarding speaker demographics or language variety

Provenance

Source

double22a via GitHub

Freshness

Updated as recently as January 2026.