Difference between revisions of "Datasets"

From rosp
m
m
Line 53: Line 53:
 
|UK English
 
|UK English
 
|{{some|1}}
 
|{{some|1}}
−
|spontaneous
+
|{{yes|spontaneous}}
 
|5
 
|5
 
|multiple dialogs
 
|multiple dialogs
Line 81: Line 81:
 
|{{n/s}}
 
|{{n/s}}
 
|{{n/s}}
 
|{{n/s}}
−
|read, spontaneous
+
|{{yes|read, spontaneous}}
 
|2
 
|2
 
|dialog
 
|dialog
Line 110: Line 110:
 
|Japanese
 
|Japanese
 
|{{dunno}}
 
|{{dunno}}
−
|spontaneous
+
|{{yes|spontaneous}}
 
|1 or 2
 
|1 or 2
 
|dialog
 
|dialog
Line 139: Line 139:
 
|US English
 
|US English
 
|{{no|0.01}}
 
|{{no|0.01}}
−
|digits
+
|{{no|digits}}
 
|1
 
|1
 
|no
 
|no
Line 168: Line 168:
 
|US English
 
|US English
 
|{{some|1}}
 
|{{some|1}}
−
|command, spontaneous
+
|{{yes|command, spontaneous}}
 
|1 or 2
 
|1 or 2
 
|no
 
|no
Line 196: Line 196:
 
|various
 
|various
 
|{{dunno}}
 
|{{dunno}}
−
|digits, command, read, spontaneous
+
|{{yes|digits, command, read, spontaneous}}
 
|1
 
|1
 
|no
 
|no
Line 225: Line 225:
 
|Japanese
 
|Japanese
 
|{{dunno}}
 
|{{dunno}}
−
|spontaneous
+
|{{yes|spontaneous}}
 
|1 to 5
 
|1 to 5
 
|meeting
 
|meeting
Line 254: Line 254:
 
|Japanese
 
|Japanese
 
|{{dunno}}
 
|{{dunno}}
−
|read
+
|{{some|read}}
 
|1
 
|1
 
|no
 
|no
Line 282: Line 282:
 
|various
 
|various
 
|{{dunno}}
 
|{{dunno}}
−
|digits, command, read, spontaneous
+
|{{yes|digits, command, read, spontaneous}}
 
|1
 
|1
 
|no
 
|no
Line 311: Line 311:
 
|US English
 
|US English
 
|{{yes|10}}
 
|{{yes|10}}
−
|read
+
|{{some|read}}
 
|1
 
|1
 
|no
 
|no
Line 339: Line 339:
 
|non-native English
 
|non-native English
 
|{{dunno}}
 
|{{dunno}}
−
|lecture
+
|{{some|lecture}}
 
|1 or more
 
|1 or more
 
|seminar
 
|seminar
Line 368: Line 368:
 
|US English
 
|US English
 
|{{no|0.01}}
 
|{{no|0.01}}
−
|digits
+
|{{no|digits}}
 
|1 or 2
 
|1 or 2
 
|full
 
|full
Line 397: Line 397:
 
|US English
 
|US English
 
|{{yes|12}}
 
|{{yes|12}}
−
|digits, command, read, dialog
+
|{{yes|digits, command, read, dialog}}
 
|1
 
|1
 
|no
 
|no
Line 426: Line 426:
 
|Japanese
 
|Japanese
 
|{{no|0.01}}
 
|{{no|0.01}}
−
|digits
+
|{{no|digits}}
 
|1
 
|1
 
|no
 
|no
Line 455: Line 455:
 
|US English, non-native English
 
|US English, non-native English
 
|{{some|1}}
 
|{{some|1}}
−
|read
+
|{{some|read}}
 
|1
 
|1
 
|no
 
|no
Line 484: Line 484:
 
|{{n/s}}
 
|{{n/s}}
 
|{{n/s}}
 
|{{n/s}}
−
|spontaneous
+
|{{yes|spontaneous}}
 
|1 to 3
 
|1 to 3
 
|full
 
|full
Line 513: Line 513:
 
|US English
 
|US English
 
|{{yes|13}}
 
|{{yes|13}}
−
|meeting
+
|{{yes|meeting}}
 
|3 to 10
 
|3 to 10
 
|meeting
 
|meeting
Line 542: Line 542:
 
|US English
 
|US English
 
|{{some|6}}
 
|{{some|6}}
−
|meeting
+
|{{yes|meeting}}
 
|3 to 9
 
|3 to 9
 
|meeting
 
|meeting
Line 571: Line 571:
 
|non-native English
 
|non-native English
 
|{{dunno}}
 
|{{dunno}}
−
|seminar, meeting
+
|{{yes|seminar, meeting}}
 
|3 to 20
 
|3 to 20
 
|seminar, meeting
 
|seminar, meeting
Line 600: Line 600:
 
|various
 
|various
 
|{{dunno}}
 
|{{dunno}}
−
|command, read, spontaneous
+
|{{yes|command, read, spontaneous}}
 
|1
 
|1
 
|no
 
|no
Line 629: Line 629:
 
|Japanese
 
|Japanese
 
|{{no|0.01}}
 
|{{no|0.01}}
−
|digits
+
|{{no|digits}}
 
|1
 
|1
 
|no
 
|no
Line 658: Line 658:
 
|Japanese
 
|Japanese
 
|{{no|0.05}}
 
|{{no|0.05}}
−
|read
+
|{{some|read}}
 
|1
 
|1
 
|no
 
|no
Line 687: Line 687:
 
|US English
 
|US English
 
|{{no|0.01}}
 
|{{no|0.01}}
−
|digits
+
|{{no|digits}}
 
|1
 
|1
 
|no
 
|no
Line 716: Line 716:
 
|UK English
 
|UK English
 
|{{some|8}}
 
|{{some|8}}
−
|meeting
+
|{{yes|meeting}}
 
|4 (18% overlap)
 
|4 (18% overlap)
 
|meeting
 
|meeting
Line 744: Line 744:
 
|UK English
 
|UK English
 
|{{no|0.05}}
 
|{{no|0.05}}
−
|command
+
|{{no|command}}
 
|2
 
|2
 
|full
 
|full
Line 773: Line 773:
 
|non-native English
 
|non-native English
 
|{{no|0.1}}
 
|{{no|0.1}}
−
|command
+
|{{no|command}}
 
|1
 
|1
 
|no
 
|no
Line 802: Line 802:
 
|US English
 
|US English
 
|{{some|2.4}}
 
|{{some|2.4}}
−
|command, dialog
+
|{{yes|command, dialog}}
 
|1 to 2
 
|1 to 2
 
|dialog
 
|dialog
Line 831: Line 831:
 
|{{n/s}}
 
|{{n/s}}
 
|{{n/s}}
 
|{{n/s}}
−
|read
+
|{{some|read}}
 
|3 or 4
 
|3 or 4
 
|full
 
|full
Line 860: Line 860:
 
|UK English
 
|UK English
 
|{{yes|10}}
 
|{{yes|10}}
−
|read
+
|{{some|read}}
 
|1 or 2
 
|1 or 2
 
|full
 
|full
Line 889: Line 889:
 
|Japanese
 
|Japanese
 
|{{no|0.01}}
 
|{{no|0.01}}
−
|digits
+
|{{no|digits}}
 
|1
 
|1
 
|no
 
|no
Line 918: Line 918:
 
|Japanese
 
|Japanese
 
|{{no|0.01}}
 
|{{no|0.01}}
−
|digits
+
|{{no|digits}}
 
|1
 
|1
 
|no
 
|no
Line 947: Line 947:
 
|Italian
 
|Italian
 
|{{dunno}}
 
|{{dunno}}
−
|command
+
|{{no|command}}
 
|4
 
|4
 
|no
 
|no
Line 975: Line 975:
 
|{{n/s}}
 
|{{n/s}}
 
|{{n/s}}
 
|{{n/s}}
−
|read
+
|{{some|read}}
 
|2
 
|2
 
|full
 
|full
Line 1,004: Line 1,004:
 
|US English, non-native English
 
|US English, non-native English
 
|{{some|5}}
 
|{{some|5}}
−
|spontaneous
+
|{{yes|spontaneous}}
 
|2 to 7
 
|2 to 7
 
|dialog
 
|dialog
Line 1,033: Line 1,033:
 
|{{n/s}}
 
|{{n/s}}
 
|{{n/s}}
 
|{{n/s}}
−
|read
+
|{{some|read}}
 
|1 or 3
 
|1 or 3
 
|full
 
|full
Line 1,061: Line 1,061:
 
|{{n/s}}
 
|{{n/s}}
 
|{{n/s}}
 
|{{n/s}}
−
|read
+
|{{some|read}}
 
|many but only 2 simultaneous
 
|many but only 2 simultaneous
 
|full
 
|full
Line 1,090: Line 1,090:
 
|UK English
 
|UK English
 
|{{no|0.05}}
 
|{{no|0.05}}
−
|command
+
|{{no|command}}
 
|1
 
|1
 
|no
 
|no
Line 1,119: Line 1,119:
 
|US English
 
|US English
 
|{{yes|11}}
 
|{{yes|11}}
−
|read
+
|{{some|read}}
 
|1
 
|1
 
|no
 
|no
Line 1,147: Line 1,147:
 
|French
 
|French
 
|{{yes|16}}
 
|{{yes|16}}
−
|spontaneous
+
|{{yes|spontaneous}}
 
|1 or more (up to 10% overlap)
 
|1 or more (up to 10% overlap)
 
|dialog
 
|dialog
Line 1,175: Line 1,175:
 
|Mandarin
 
|Mandarin
 
|{{dunno}}
 
|{{dunno}}
−
|spontaneous
+
|{{yes|spontaneous}}
 
|1 or more
 
|1 or more
 
|dialog
 
|dialog
Line 1,203: Line 1,203:
 
|Arabic
 
|Arabic
 
|{{dunno}}
 
|{{dunno}}
−
|spontaneous
+
|{{yes|spontaneous}}
 
|1 or more
 
|1 or more
 
|dialog
 
|dialog
Line 1,232: Line 1,232:
 
|UK English
 
|UK English
 
|{{yes|10}}
 
|{{yes|10}}
−
|read
+
|{{some|read}}
 
|1
 
|1
 
|no
 
|no
Line 1,261: Line 1,261:
 
|various
 
|various
 
|{{dunno}}
 
|{{dunno}}
−
|various
+
|{{yes|command, read, spontaneous}}
 
|1 or more
 
|1 or more
 
|simulated
 
|simulated

Revision as of 22:06, 8 August 2014

This page aims to provide a list of datasets with detailed attributes and links to corresponding research results (papers, numerical results, output transcriptions, intermediary data, etc). Each dataset may be used for one or more applications: automatic speech recognition, speaker identification and verification, source localization, speech enhancement and separation...

Disclaimer: Only publicly available datasets with a total duration longer than 5 min are listed.

Datasets General attributes Speech Channel Noise Ground truth
year use case total time (h) sam. rate (kHz) dist. / noisy mics video cams cost links speak. time (h) uniq. speak. lang. uniq. words (k) speak. style speak. overl. overl. type chan. type speak. radiat. speak. loc. speak. moves noise type ref. signal speak. loc., orient. words non- verb. traits noise events
ShATR 1994 meeting 0.6 48 3 no free download

email paper

0.6 5 UK English 1 spontaneous 5 multiple dialogs reverb human quasi-fixed head meeting headset yes yes no yes
LLSEC 1996 dialog 1.4 16 4 no free download

email

? 12 N/S N/S read, spontaneous 2 dialog reverb human quasi-fixed head hallway, restaurant (scenarized) no yes no no no
RWCP Spoken Dialog Corpus 1996 - 1997 dialog 10 16 2 no free download

email paper

10 39 Japanese ? spontaneous 1 or 2 dialog reverb (low) human quasi-fixed head stationary background no no yes no no
Aurora-2 2000 public spaces 33 8 - 16 1 no free given TIDigits download

email paper

33 214 US English 0.01 digits 1 no simulated phone human N/S no various real environments original N/S yes no yes
SPINE1, SPINE2 2000 - 2001 military 38 16 2 no 7.4 k$ purchase

email paper

? 100 US English 1 command, spontaneous 1 or 2 no simulated radio human quasi-fixed head military no no yes no no
Aurora-3 (subset of SpeechDat- Car) 2000 - 2003 car ? 16 4 no 1 k€ purchase

papers

? ? various ? digits, command, read, spontaneous 1 no reverb human quasi-fixed head car headset no yes no no
RWCP Meeting Speech Corpus 2001 meeting 3.5 16 - 48 1 3 free download

email paper

3.5 ? Japanese ? spontaneous 1 to 5 meeting reverb (low) human quasi-fixed head stationary background headset no yes no no
RWCP Real Environment Speech and Acoustic Database 2001 domestic, office ? 16 - 48 30 no free download

email paper

? 5 Japanese ? read 1 no real rir, reverb loudspeaker various no, pivoting arm stationary background original yes yes no yes
SpeechDat- Car 2001 - 2011 car ? 16 4 no 39 - 182 k€ per lang purchase

paper

? 300 per lang various ? digits, command, read, spontaneous 1 no reverb human quasi-fixed head car headset no yes no no
Aurora-4 2002 public spaces ? 8 - 16 1 no free given WSJ0 download

email paper

? 101 US English 10 read 1 no simulated phone human N/S no various real environments original N/S yes no yes
TED 2002 seminar 47 16 1 no 0.5 k$ purchase

paper

47 188 non-native English ? lecture 1 or more seminar reverb human quasi-fixed head stationary background lapel no partial no no
CUAVE 2002 cocktail party 3 44 1 1 free download

email paper

3 36 US English 0.01 digits 1 or 2 full reverb human quasi-fixed head stationary background no no yes no no
CU-Move Microphone Array Data 2002 - 2011 car 286 44 6 - 8 no 25 k$ purchase

email paper

286 172 US English 12 digits, command, read, dialog 1 no reverb human quasi-fixed head car no no yes no no
CENSREC-1 (Aurora-2J) 2003 public spaces ? 8 1 no free download

email paper

? 214 Japanese 0.01 digits 1 no simulated phone human N/S no various real environments original N/S yes no yes
AVICAR 2004 car 29 16 7 4 free download

email paper

29 86 US English, non-native English 1 read 1 no reverb human quasi-fixed head car no no yes no no
AV16.3 2004 meeting 1.5 16 16 3 free download

email paper

1.5 12 N/S N/S spontaneous 1 to 3 full reverb human various walk stationary background no yes no no no
ICSI Meeting Corpus 2004 meeting 72 16 6 no 2.8 k$ purchase

email paper

72 53 US English 13 meeting 3 to 10 meeting reverb human quasi-fixed head stationary background headset, lapel no yes yes no
NIST Meeting Pilot Corpus Speech 2004 meeting 15 16 7 no 5.5 k$ purchase

email paper

15 61 US English 6 meeting 3 to 9 meeting reverb human various walk stationary background headset, lapel no yes no no
CHIL Meetings 2004 - 2007 seminar, meeting 60 44 79 - 147 6 - 9 3.5 k€ purchase

email paper

? ? non-native English ? seminar, meeting 3 to 20 seminar, meeting reverb human quasi-fixed head meeting (scenarized) headset yes yes yes no
SPEECON 2004 - 2011 public space, domestic, office, car ? 16 3 no 75 k€ per lang purchase

email paper

? 600 per lang various ? command, read, spontaneous 1 no reverb human quasi-fixed head various real environments headset no yes no no
CENSREC-2 2005 car ? 16 1 no free download

email paper

? 214 Japanese 0.01 digits 1 no reverb human quasi-fixed head car headset no yes no no
CENSREC-3 2005 car ? 16 1 no 21 k¥ purchase

email paper

? 311 Japanese 0.05 read 1 no reverb human quasi-fixed head car headset no yes no no
Aurora-5 2006 public spaces, domestic, office, car ? 8 1 no free given TIDigits download

email paper

? 225 US English 0.01 digits 1 no no, simulated rir, real rir loudspeaker N/S no various real environments original no yes no yes
AMI 2006 meeting 100 16 16 6 free download

email paper

? 189 UK English 8 meeting 4 (18% overlap) meeting reverb human quasi-fixed head stationary background headset, lapel yes yes yes no
PASCAL SSC 2006 cocktail party 8.8 25 1 no free email

paper

8.8 34 UK English 0.05 command 2 full no human N/S no no original N/S yes no no
HIWIRE 2007 airplane 21 16 1 no 0.05 k€ purchase

email paper

21 81 non-native English 0.1 command 1 no no human N/S head airplane original N/S yes no no
UT-Drive 2007 car 40 25 5 2 25 k$ download

email paper

40 25 US English 2.4 command, dialog 1 to 2 dialog reverb human quasi-fixed head car headset (low quality) no partial no no
SASSEC, SiSEC under- determined 2007 - 2011 cocktail party 0.3 16 2 no free download

email paper

0.3 16 N/S N/S read 3 or 4 full simulated rir, real rir, reverb no, loudspeaker fixed no no original, spatial image yes no no no
MC-WSJ-AV, PASCAL SSC2, 2012_MMA, REVERB RealData 2007 - 2014 cocktail party 10 16 8 - 40 no 1.5 k$ purchase

email paper paper

? 45 UK English 10 read 1 or 2 full reverb human various walk stationary background headset, lapel yes yes no no
CENSREC-4 (Simulated) 2008 public spaces, domestic, office, car ? 16 1 no free download

email paper

? 214 Japanese 0.01 digits 1 no real rir dummy fixed no various real environments original no yes no yes
CENSREC-4 (Real) 2008 public spaces, domestic, office, car ? 16 1 no free download

email paper

? 10 Japanese 0.01 digits 1 no reverb human quasi-fixed head various real environments headset no yes no yes
DICIT 2008 domestic 6 48 16 2 free download

email paper

1 ? Italian ? command 4 no reverb human various walk domestic (scenarized) headset, tv yes yes no yes
SiSEC head-geometry 2008 cocktail party 1.9 16 2 no free download email

paper

1.9 ? N/S N/S read 2 full real rir loudspeaker various no no original, spatial image yes no no no
COSINE 2009 dialog 38 48 20 no free download

email paper

11 91 US English, non-native English 5 spontaneous 2 to 7 dialog reverb human various walk various real environments headset, throat mic no yes no no
SiSEC real-world noise 2010 public spaces 0.3 16 2 - 4 no free download

email paper

0.3 6 N/S N/S read 1 or 3 full no, reverb (other room) loudspeaker various no various real environments original, spatial image yes no no no
SiSEC dynamic 2010 - 2011 cocktail party 0.2 16 2 - 4 no free download email

paper

0.2 ? N/S N/S read many but only 2 simultaneous full reverb loudspeaker various simulated no original, spatial image yes no no no
CHiME 1, CHiME 2 Grid 2011 - 2012 domestic 70 16 - 48 2 no free download

email paper

12 34 UK English 0.05 command 1 no real rir dummy quasi-fixed simulated head domestic yes yes yes no no
CHiME 2 WSJ0 2012 domestic 78 16 2 no free given WSJ0 download

email paper

33 101 US English 11 read 1 no real rir dummy fixed no domestic yes yes yes no no
ETAPE 2012 TV/radio debates, outdoor interviews 42 16 1 1 ? email

paper

32 347 French 16 spontaneous 1 or more (up to 10% overlap) dialog reverb (some) human quasi-fixed head various real environments no N/S yes no yes
GALE (Chinese broadcast conversation) 2013 TV dialog 120 16 1 no 3.5 k$ purchase

email

108 ? Mandarin ? spontaneous 1 or more dialog no human quasi-fixed head no no N/S yes no no
GALE (Arabic broadcast conversation) 2013 TV dialog 251 16 1 no 7 k$ purchase

email

234 ? Arabic ? spontaneous 1 or more dialog no human quasi-fixed head no no N/S yes no no
REVERB SimData 2013 domestic, office 25 16 8 no free given WSJCAM0 purchase

email paper

25 130 UK English 10 read 1 no real rir loudspeaker fixed no stationary background original, spatial image yes yes no yes
DIRHA 2014 domestic 3.8 48 40 no free download

email paper

1.3 30 various ? command, read, spontaneous 1 or more simulated real rir loudspeaker various no domestic (sum of events) yes yes yes no yes

Automatic speech recognition

1st CHiME Challenge (2011)

Artificially distorted version of the small vocabulary GRID audio-visual corpus (audio only). Binaural reverberated speech with speaker situated in front of the microphones. Additive household noises impinging from different directions. Clean-training, noisy-training, development and evaluation sets available, see

Jon Barker, E. Vincent, N. Ma, H. Christensen, P. Green, "The PASCAL CHiME speech separation and recognition challenge", Computer Speech & Language, Volume 27, Issue 3, May 2013, Pages 621-633.

Available from Computer Speech and Language here

Corpus available here (no cost)

Resources

  • Training recipe of the challenge for HTK here.

Baselines

  • See the paper above for results for a wide range of techniques.


AURORA 5 (2007)

Artificially distorted version of the digits TI-DIGITS corpus. Additive noise and additive noise plus reverberant speech sets. Variable SNR range. Various mixed training sets, no evaluation set, see

G. Hirsch "Aurora-5 Experimental Framework for the Performance Evaluation of Speech Recognition in Case of a Hands-free Speech Input in Noisy Environments", Niederrhein University of Applied Sciences, 2007.

Paper available online here (no cost)

Corpus available from LDC here

Resources

  • Training recipe for HTK is provided with the corpora.

Baselines

  • Reproducible baseline: The above cited paper includes a baseline for the ETSI Advanced Front-End.


AURORA 4 (2002)

Artificially distorted version of the 5K word Wall Street Journal corpus (WSJ0). Stationary and non-stationary noises added. Second recordings with distant mismatched microphone. Clean-training, mixed-training, noisy training and test sets available. No evaluation set, see

G. Hirsch "Experimental Framework for the Performance Evaluation of Speech Recognition Front-ends on a Large Vocabulary Task", ETSI STQ Aurora DSR Working Group, 2002.

Paper available with the corpus.

Corpora available from ELRA here and here

Resources

  • Training recipe for HTK available here. Note that this recipe is for Wall-Street Journal (WSJ0), which is the clean speech version of AURORA4. Small changes are needed in the feature extraction scripts to account for different file terminations.

Speaker identification and verification

Speech enhancement and separation

Other applications

Contribute a dataset

To contribute a new dataset, please

  • create an account and login
  • go to the wiki page above corresponding to your application; if it does not exist yet, you may create it
  • click on the "Edit" link at the top of the page and add a new section for your dataset (the datasets are ordered by year of collection)
  • click on the "Save page" link at the bottom of the page to save your modifications

Please make sure to provide the following information:

  • name of the dataset and year of collection
  • authors, institution, contact information
  • link to the dataset and to side resources (lexicon, language model, etc)
  • short description (nature of the data, license, etc) and link to a paper/report describing the dataset, if any
  • at least 1 research result obtained for this dataset (see below)

We currently cannot provide storage space for large datasets. Please upload the dataset at a stable URL on the website of your institution or elsewhere and provide its URL only. If this is not possible, please contact the resources sharing working group.

Contribute a research result

To contribute a new research result, please

  • create an account and login
  • go to the wiki page and the section corresponding to the dataset for which this result was obtained
  • click on the "Edit" link on the right of the section header and add a new item for your result
  • click on the "Save page" link at the bottom of the page to save your modifications

Please make sure to provide the following information:

  • authors, paper/report title, means of publication
  • link to the pdf of the paper
  • link to derived data (output transcriptions, intermediary data, etc)
  • Code and instructions to reproduce experiments (if available)

In order to save storage space, please do not upload the paper on this wiki, but link it as much as possible from your institutional archive, from another public archive (e.g., arxiv) or from the publisher website (e.g., ieexplore).

We currently cannot provide storage space for large datasets. Please upload the derived data at a stable URL on the website of your institution or elsewhere and provide its URL only. If this is not possible, please contact the resources sharing working group.

Retrieved from "https://wiki.inria.fr/wikis/rosp/index.php?title=Datasets&oldid=191"