CSR-I (WSJ0) Complete
| Item Name: | CSR-I (WSJ0) Complete |
| Author(s): | John S. Garofolo, David Graff, Janet M. Baker, Doug Paul, David Pallett |
| LDC Catalog No.: | LDC93S6A |
| ISBN: | 1-58563-006-3 |
| ISLRN: | 296-840-353-630-9 |
| DOI: | https://doi.org/10.35111/ewkm-cg47 |
| Release Date: | May 30, 2007 |
| Member Year(s): | 1993, 1996 |
| DCMI Type(s): | Sound |
| Sample Type: | 1-channel pcm compressed |
| Sample Rate: | 16000 |
| Data Source(s): | microphone speech |
| Project(s): | DARPA-CSR |
| Application(s): | speech recognition |
| Language(s): | English |
| Language ID(s): | eng |
| License(s): |
LDC User Agreement for Non-Members |
| Online Documentation: | LDC93S6A Documents |
| Licensing Instructions: | Subscription & Standard Members, and Non-Members |
| Citation: | Garofolo, John S., et al. CSR-I (WSJ0) Complete LDC93S6A. Web Download. Philadelphia: Linguistic Data Consortium, 1993. |
| Related Works: | View |
Introduction
CSR-I (WSJ0) Complete was developed by NIST and contains approximately 141 hours of speech recordings of 123 speakers reading excerpts from the Wall Street Journal. About half the speakers are male and half female. Additionally, the discs contain complete orthographic transcriptions of the speech data and complete bigram language models for the Wall Street Journal text data from which the prompting text was taken. These materials are provided on disc 11-4 and also on disc 11-10. Disc 11-4 also contains the complete text of the WSJ articles from which the utterance prompts and language models were derived.
During 1991, the DARPA Spoken Language Program initiated efforts to build a new corpus to support research on large-vocabulary Continuous Speech Recognition (CSR) systems.
The first two CSR Corpora consist primarily of read speech with texts drawn from a machine-readable corpus of Wall Street Journal news text and are thus often known as WSJ0 and WSJ1. (Later sections of the CSR set of corpora, however, will consist of read texts from other sources of North American business news and eventually from other news domains).
The texts to be read were selected to fall within either a 5,000-word or a 20,000-word subset of the WSJ text corpus. (See the documentation for details). Some spontaneous dictation is included in addition to the read speech. The dictation portion was collected using journalists who dictated hypothetical news articles.
Data
The data was recorded at SRI, TI and MIT. Two microphones are used throughout: a close-talking Sennheiser HMD414 and a secondary microphone, which may vary. These two channels have been stored into separate files and organized in a parallel structure in the directories; six of the discs (11-1 through 11-6) contain the primary channel, and the other six contain the secondary channel. Header information in each waveform file provides details on the recording source and conditions.
Audio files are presented as single channel, 16-bit, 16 kHz sphere files.
Samples
Please listen to this audio sample.
Updates
None at this time.