Open Speech and Language Resources



LibriSpeech ASR corpus

Identifier: SLR12

Summary: Large-scale (1000 hours) corpus of read English speech

Category: Speech

License: CC BY 4.0

Downloads (use a mirror closer to you):
dev-clean.tar.gz [error getting size]   (development set, "clean" speech )   Mirrors: [US]   [EU]   [CN]  
dev-other.tar.gz [error getting size]   (development set, "other", more challenging, speech )   Mirrors: [US]   [EU]   [CN]  
test-clean.tar.gz [error getting size]   (test set, "clean" speech )   Mirrors: [US]   [EU]   [CN]  
test-other.tar.gz [error getting size]   (test set, "other" speech )   Mirrors: [US]   [EU]   [CN]  
train-clean-100.tar.gz [error getting size]   (training set of 100 hours "clean" speech )   Mirrors: [US]   [EU]   [CN]  
train-clean-360.tar.gz [error getting size]   (training set of 360 hours "clean" speech )   Mirrors: [US]   [EU]   [CN]  
train-other-500.tar.gz [error getting size]   (training set of 500 hours "other" speech )   Mirrors: [US]   [EU]   [CN]  
intro-disclaimers.tar.gz [error getting size]   (extracted LibriVox announcements for some of the speakers )   Mirrors: [US]   [EU]   [CN]  
original-mp3.tar.gz [error getting size]   (LibriVox mp3 files, from which corpus' audio was extracted )   Mirrors: [US]   [EU]   [CN]  
original-books.tar.gz [error getting size]   (Project Gutenberg texts, against which the audio in the corpus was aligned )   Mirrors: [US]   [EU]   [CN]  
raw-metadata.tar.gz [error getting size]   (Some extra meta-data produced during the creation of the corpus )   Mirrors: [US]   [EU]   [CN]  
md5sum.txt [error getting size]   (MD5 checksums for the archive files )   Mirrors: [US]   [EU]   [CN]  

About this resource:

LibriSpeech is a corpus of approximately 1000 hours of 16kHz read English speech, 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.

Acoustic models, trained on this data set, are available at kaldi-asr.org and language models, suitable for evaluation can be found at http://www.openslr.org/11/.

For more information, see the paper "LibriSpeech: an ASR corpus based on public domain audio books", Vassil Panayotov, Guoguo Chen, Daniel Povey and Sanjeev Khudanpur, ICASSP 2015 (submitted) (pdf)