• Description:

LibriSpeech is a corpus of approximately 1000 hours of read English speech with sampling rate of 16 kHz, prepared by Vassil Panayotov with the assistance of Daniel Povey. The data is derived from read audiobooks from the LibriVox project, and has been carefully segmented and aligned.

Split Examples
'dev_clean' 2,703
'dev_other' 2,864
'test_clean' 2,620
'test_other' 2,939
'train_clean100' 28,539
'train_clean360' 104,014
'train_other500' 148,688
  • Feature structure:
    'chapter_id': tf.int64,
    'id': tf.string,
    'speaker_id': tf.int64,
    'speech': Audio(shape=(None,), dtype=tf.int16),
    'text': Text(shape=(), dtype=tf.string),
  • Feature documentation:
Feature Class Shape Dtype Description
chapter_id Tensor tf.int64
id Tensor tf.string
speaker_id Tensor tf.int64
speech Audio (None,) tf.int16
text Text tf.string
  • Citation:
  title={Librispeech: an ASR corpus based on public domain audio books},
  author={Panayotov, Vassil and Chen, Guoguo and Povey, Daniel and Khudanpur, Sanjeev},
  booktitle={Acoustics, Speech and Signal Processing (ICASSP), 2015 IEEE International Conference on},