ASR

LibriSpeech is a widely used dataset for training and evaluating automatic speech recognition systems, derived from audiobooks.