user_libri_audio

  • Description:

UserLibri is a dataset containing paired audio-transcripts and additional text only data for each of 107 users. It is a reformatting of the LibriSpeech dataset found at http://www.openslr.org/12, reorganizing the data into users with an average of 52 LibriSpeech utterances and about 6,700 text example sentences per user. The UserLibriAudio class provides access to the audio-transcript pairs. See UserLibriText for the additional text data.

Split Examples
'test-clean_speaker-1089-book-4217' 64
'test-clean_speaker-1188-book-20019' 45
'test-clean_speaker-121-book-1041' 10
'test-clean_speaker-121-book-1989' 15
'test-clean_speaker-121-book-209' 37
'test-clean_speaker-1221-book-33' 41
'test-clean_speaker-1284-book-732' 8
'test-clean_speaker-1284-book-955' 55
'test-clean_speaker-1320-book-940' 59
'test-clean_speaker-1580-book-108' 105
'test-clean_speaker-1995-book-15265' 72
'test-clean_speaker-2094-book-507' 61
'test-clean_speaker-2300-book-820' 42
'test-clean_speaker-237-book-24' 62
'test-clean_speaker-237-book-2770' 26
'test-clean_speaker-260-book-11' 21
'test-clean_speaker-260-book-3748' 61
'test-clean_speaker-2830-book-1549' 90
'test-clean_speaker-2961-book-1572' 46
'test-clean_speaker-3570-book-833' 50
'test-clean_speaker-3575-book-1827' 57
'test-clean_speaker-3729-book-2981' 47
'test-clean_speaker-4077-book-5630' 39
'test-clean_speaker-4446-book-94' 108
'test-clean_speaker-4507-book-135' 60
'test-clean_speaker-4970-book-3178' 63
'test-clean_speaker-4992-book-10540' 41
'test-clean_speaker-4992-book-22002' 21
'test-clean_speaker-5105-book-1353' 56
'test-clean_speaker-5142-book-2300' 7
'test-clean_speaker-5142-book-24811' 69
'test-clean_speaker-5142-book-7891' 26
'test-clean_speaker-5639-book-14420' 42
'test-clean_speaker-5683-book-9983' 75
'test-clean_speaker-61-book-28700' 104
'test-clean_speaker-672-book-1597' 75
'test-clean_speaker-6829-book-13110' 91
'test-clean_speaker-6930-book-13158' 28
'test-clean_speaker-6930-book-2681' 21
'test-clean_speaker-6930-book-30905' 29
'test-clean_speaker-7021-book-11667' 31
'test-clean_speaker-7021-book-26177' 28
'test-clean_speaker-7127-book-2681' 71
'test-clean_speaker-7176-book-13441' 46
'test-clean_speaker-7176-book-38675' 28
'test-clean_speaker-7729-book-6812' 47
'test-clean_speaker-8224-book-19215' 32
'test-clean_speaker-8230-book-2529' 44
'test-clean_speaker-8455-book-27067' 71
'test-clean_speaker-8463-book-15263' 15
'test-clean_speaker-8463-book-2488' 59
'test-clean_speaker-8555-book-36508' 16
'test-clean_speaker-8555-book-39159' 46
'test-clean_speaker-908-book-2002' 26
'test-clean_speaker-908-book-574' 31
'test-other_speaker-1688-book-4276' 96
'test-other_speaker-1998-book-19019' 28
'test-other_speaker-1998-book-28725' 87
'test-other_speaker-2033-book-3436' 52
'test-other_speaker-2414-book-1998' 60
'test-other_speaker-2414-book-26379' 2
'test-other_speaker-2414-book-40359' 36
'test-other_speaker-2609-book-12434' 39
'test-other_speaker-2609-book-18096' 15
'test-other_speaker-2609-book-27090' 25
'test-other_speaker-3005-book-76' 108
'test-other_speaker-3080-book-12544' 61
'test-other_speaker-3331-book-2787' 73
'test-other_speaker-3528-book-135' 144
'test-other_speaker-3538-book-10136' 27
'test-other_speaker-3538-book-540' 67
'test-other_speaker-367-book-5921' 21
'test-other_speaker-367-book-9464' 34
'test-other_speaker-3764-book-135' 113
'test-other_speaker-3997-book-14958' 21
'test-other_speaker-3997-book-1608' 66
'test-other_speaker-4198-book-16653' 31
'test-other_speaker-4198-book-8166' 60
'test-other_speaker-4294-book-12176' 27
'test-other_speaker-4294-book-135' 30
'test-other_speaker-4294-book-4028' 19
'test-other_speaker-4294-book-9983' 6
'test-other_speaker-4350-book-1399' 34
'test-other_speaker-4350-book-4602' 61
'test-other_speaker-4852-book-28952' 113
'test-other_speaker-533-book-434' 25
'test-other_speaker-533-book-969' 71
'test-other_speaker-5442-book-1399' 60
'test-other_speaker-5442-book-9983' 20
'test-other_speaker-5484-book-5516' 72
'test-other_speaker-5764-book-38804' 98
'test-other_speaker-6070-book-1184' 50
'test-other_speaker-6070-book-33800' 19
'test-other_speaker-6128-book-19717' 68
'test-other_speaker-6432-book-16127' 118
'test-other_speaker-6938-book-3076' 31
'test-other_speaker-7018-book-3440' 52
'test-other_speaker-7105-book-1477' 80
'test-other_speaker-7902-book-21297' 135
'test-other_speaker-7975-book-24585' 100
'test-other_speaker-8131-book-20212' 118
'test-other_speaker-8188-book-19215' 12
'test-other_speaker-8188-book-41326' 116
'test-other_speaker-8280-book-14566' 66
'test-other_speaker-8461-book-3441' 17
'test-other_speaker-8461-book-6328' 39
'test-other_speaker-8461-book-9189' 16
  • Feature structure:
FeaturesDict({
    'audio': Audio(shape=(None,), dtype=tf.int64),
    'book_id': Text(shape=(), dtype=tf.string),
    'id': Text(shape=(), dtype=tf.string),
    'speaker_id': Text(shape=(), dtype=tf.string),
    'transcript': Text(shape=(), dtype=tf.string),
    'user_id': Text(shape=(), dtype=tf.string),
})
  • Feature documentation:
Feature Class Shape Dtype Description
FeaturesDict
audio Audio (None,) tf.int64 The audio clip containing a snippet from a book read aloud
book_id Text tf.string The book that this utterance is read from
id Text tf.string The id of this unique utterance
speaker_id Text tf.string The speaker who read this utterance
transcript Text tf.string The text that the speaker read to produce the audio
user_id Text tf.string The user that this utterance belongs to (unique speaker and book)
  • Citation:
@inproceedings{breiner2022userlibri,
  title={UserLibri: A Dataset for ASR Personalization Using Only Text},
  author={Breiner, Theresa and Ramaswamy, Swaroop and Variani, Ehsan and Garg, Shefali and Mathews, Rajiv and Sim, Khe Chai and Gupta, Kilol and Chen, Mingqing and McConnaughey, Lara},
  booktitle={Proc. Interspeech 2022},
  year={2022}
}