TasksSotADatasetsPapersMethodsSubmitAbout
Papers With Code 2

A community resource for machine learning research: papers, code, benchmarks, and state-of-the-art results.

Explore

Notable BenchmarksAll SotADatasetsPapersMethods

Community

Submit ResultsAbout

Data sourced from the PWC Archive (CC-BY-SA 4.0). Built by the community, for the community.

Models/Deep Speech 2

Deep Speech 2

Reported on 14 benchmarks across 2 tasks · 1 paper · 13 SOTA

Note: results are matched by exact model name. Different papers may use the same name for different model variants.

Audio14 results

  • Speech RecognitiononWSJ eval93
    Word Error Rate (WER)· uses extra data· 2015-12-08
    4.98
    SOTA
    Deep Speech 2: End-to-End Speech Recognition in English and MandarinarXiv:1512.02595
  • Speech RecognitiononLibriSpeech test-clean
    Word Error Rate (WER)· 2015-12-08
    5.33
    best: 0.985 (United Med ASR)
    SOTA
    Deep Speech 2: End-to-End Speech Recognition in English and MandarinarXiv:1512.02595
  • Speech RecognitiononLibriSpeech test-other
    Word Error Rate (WER)· 2015-12-08
    13.25
    best: 2.48 (SAMBA ASR)
    SOTA
    Deep Speech 2: End-to-End Speech Recognition in English and MandarinarXiv:1512.02595
  • Speech RecognitiononVoxForge European
    Percentage error· 2015-12-08
    17.55
    SOTA
    Deep Speech 2: End-to-End Speech Recognition in English and MandarinarXiv:1512.02595
  • Speech RecognitiononVoxForge American-Canadian
    Percentage error· 2015-12-08
    7.55
    SOTA
    Deep Speech 2: End-to-End Speech Recognition in English and MandarinarXiv:1512.02595
  • Speech RecognitiononVoxForge Indian
    Percentage error· 2015-12-08
    22.44
    SOTA
    Deep Speech 2: End-to-End Speech Recognition in English and MandarinarXiv:1512.02595
  • Speech RecognitiononVoxForge Commonwealth
    Percentage error· 2015-12-08
    13.56
    SOTA
    Deep Speech 2: End-to-End Speech Recognition in English and MandarinarXiv:1512.02595
  • Speech RecognitiononCHiME real
    Percentage error· 2015-12-08
    21.79
    best: 7.8 (Distortion-Independent + WRBN + Utterance-Wise Recurrent Dropout + Magnitude Features)
    SOTA
    Deep Speech 2: End-to-End Speech Recognition in English and MandarinarXiv:1512.02595
  • Speech RecognitiononCHiME clean
    Percentage error· 2015-12-08
    3.34
    SOTA
    Deep Speech 2: End-to-End Speech Recognition in English and MandarinarXiv:1512.02595
  • Accented Speech RecognitiononVoxForge European
    Percentage error· 2015-12-08
    17.55
    SOTA
    Deep Speech 2: End-to-End Speech Recognition in English and MandarinarXiv:1512.02595
  • Accented Speech RecognitiononVoxForge American-Canadian
    Percentage error· 2015-12-08
    7.55
    SOTA
    Deep Speech 2: End-to-End Speech Recognition in English and MandarinarXiv:1512.02595
  • Accented Speech RecognitiononVoxForge Indian
    Percentage error· 2015-12-08
    22.44
    SOTA
    Deep Speech 2: End-to-End Speech Recognition in English and MandarinarXiv:1512.02595
  • Accented Speech RecognitiononVoxForge Commonwealth
    Percentage error· 2015-12-08
    13.56
    SOTA
    Deep Speech 2: End-to-End Speech Recognition in English and MandarinarXiv:1512.02595
  • Speech RecognitiononWSJ eval92
    Word Error Rate (WER)· uses extra data· 2015-12-08
    3.6
    best: 1.3 (Speechstew 100M)
    Deep Speech 2: End-to-End Speech Recognition in English and MandarinarXiv:1512.02595