[
  {
    "Word Error Rate (WER)": 4.11,
    "code_links": [
      {
        "title": "facebookresearch/wav2letter",
        "url": "https://github.com/facebookresearch/wav2letter/tree/master/recipes/models/sota/2019"
      }
    ],
    "date": "2019-11-19",
    "date2": 20191119,
    "model": "Conv + Transformer AM (ConvLM  with Transformer Rescoring)",
    "paper": {
      "title": "End-to-end ASR: from Supervised to Semi-Supervised Learning with Modern Architectures",
      "url": "https://cknow.io/lib/25e0202f87d40fc8"
    },
    "paper_data_uoa": "25e0202f87d40fc8"
  },
  {
    "Word Error Rate (WER)": 5.18,
    "code_links": [
      {
        "title": "facebookresearch/wav2letter",
        "url": "https://github.com/facebookresearch/wav2letter/tree/master/recipes/models/sota/2019"
      }
    ],
    "date": "2019-11-19",
    "date2": 20191119,
    "model": "Conv + Transformer AM (ConvLM  with Transformer Rescoring) (LS only)",
    "paper": {
      "title": "End-to-end ASR: from Supervised to Semi-Supervised Learning with Modern Architectures",
      "url": "https://cknow.io/lib/25e0202f87d40fc8"
    },
    "paper_data_uoa": "25e0202f87d40fc8"
  },
  {
    "Word Error Rate (WER)": 5.8,
    "code_links": [
      {
        "title": "mozilla/DeepSpeech",
        "url": "https://github.com/mozilla/DeepSpeech"
      },
      {
        "title": "makcedward/nlpaug",
        "url": "https://github.com/makcedward/nlpaug"
      },
      {
        "title": "shelling203/SpecAugment",
        "url": "https://github.com/shelling203/SpecAugment"
      },
      {
        "title": "DemisEom/SpecAugment",
        "url": "https://github.com/DemisEom/SpecAugment"
      },
      {
        "title": "lRomul/argus-freesound",
        "url": "https://github.com/lRomul/argus-freesound"
      },
      {
        "title": "ZhengkunTian/OpenTransformer",
        "url": "https://github.com/ZhengkunTian/OpenTransformer"
      },
      {
        "title": "ebouteillon/freesound-audio-tagging-2019",
        "url": "https://github.com/ebouteillon/freesound-audio-tagging-2019"
      },
      {
        "title": "sh951011/Korean-Speech-Recognition",
        "url": "https://github.com/sh951011/Korean-Speech-Recognition/blob/master/package/feature.py"
      },
      {
        "title": "KimJeongSun/SpecAugment_numpy_scipy",
        "url": "https://github.com/KimJeongSun/SpecAugment_numpy_scipy"
      },
      {
        "title": "hgstudent/las",
        "url": "https://github.com/hgstudent/las"
      },
      {
        "title": "viig99/mixmatch-freesound",
        "url": "https://github.com/viig99/mixmatch-freesound"
      },
      {
        "title": "knlee-voice/PaperNotes",
        "url": "https://github.com/knlee-voice/PaperNotes"
      }
    ],
    "date": "2019-04-18",
    "date2": 20190418,
    "model": "LAS + SpecAugment",
    "paper": {
      "title": "SpecAugment: A Simple Data Augmentation Method for Automatic Speech Recognition",
      "url": "https://cknow.io/lib/15408ea14d4c82f9"
    },
    "paper_data_uoa": "15408ea14d4c82f9"
  },
  {
    "Word Error Rate (WER)": 6.5,
    "code_links": [
      {
        "title": "mozilla/DeepSpeech",
        "url": "https://github.com/mozilla/DeepSpeech"
      },
      {
        "title": "makcedward/nlpaug",
        "url": "https://github.com/makcedward/nlpaug"
      },
      {
        "title": "shelling203/SpecAugment",
        "url": "https://github.com/shelling203/SpecAugment"
      },
      {
        "title": "DemisEom/SpecAugment",
        "url": "https://github.com/DemisEom/SpecAugment"
      },
      {
        "title": "lRomul/argus-freesound",
        "url": "https://github.com/lRomul/argus-freesound"
      },
      {
        "title": "ZhengkunTian/OpenTransformer",
        "url": "https://github.com/ZhengkunTian/OpenTransformer"
      },
      {
        "title": "ebouteillon/freesound-audio-tagging-2019",
        "url": "https://github.com/ebouteillon/freesound-audio-tagging-2019"
      },
      {
        "title": "sh951011/Korean-Speech-Recognition",
        "url": "https://github.com/sh951011/Korean-Speech-Recognition/blob/master/package/feature.py"
      },
      {
        "title": "KimJeongSun/SpecAugment_numpy_scipy",
        "url": "https://github.com/KimJeongSun/SpecAugment_numpy_scipy"
      },
      {
        "title": "hgstudent/las",
        "url": "https://github.com/hgstudent/las"
      },
      {
        "title": "viig99/mixmatch-freesound",
        "url": "https://github.com/viig99/mixmatch-freesound"
      },
      {
        "title": "knlee-voice/PaperNotes",
        "url": "https://github.com/knlee-voice/PaperNotes"
      }
    ],
    "date": "2019-04-18",
    "date2": 20190418,
    "model": "LAS (no LM)",
    "paper": {
      "title": "SpecAugment: A Simple Data Augmentation Method for Automatic Speech Recognition",
      "url": "https://cknow.io/lib/15408ea14d4c82f9"
    },
    "paper_data_uoa": "15408ea14d4c82f9"
  },
  {
    "Word Error Rate (WER)": 7.84,
    "code_links": [
      {
        "title": "NVIDIA/OpenSeq2Seq",
        "url": "https://github.com/NVIDIA/OpenSeq2Seq"
      },
      {
        "title": "stefanpantic/asr",
        "url": "https://github.com/stefanpantic/asr"
      }
    ],
    "date": "2019-04-05",
    "date2": 20190405,
    "model": "Jasper DR 10x5 (+ Time/Freq Masks)",
    "paper": {
      "title": "Jasper: An End-to-End Convolutional Neural Acoustic Model",
      "url": "https://cknow.io/lib/d52115ec6c741b09"
    },
    "paper_data_uoa": "d52115ec6c741b09"
  },
  {
    "Word Error Rate (WER)": 8.79,
    "code_links": [
      {
        "title": "NVIDIA/OpenSeq2Seq",
        "url": "https://github.com/NVIDIA/OpenSeq2Seq"
      },
      {
        "title": "stefanpantic/asr",
        "url": "https://github.com/stefanpantic/asr"
      }
    ],
    "date": "2019-04-05",
    "date2": 20190405,
    "model": "Jasper DR 10x5",
    "paper": {
      "title": "Jasper: An End-to-End Convolutional Neural Acoustic Model",
      "url": "https://cknow.io/lib/d52115ec6c741b09"
    },
    "paper_data_uoa": "d52115ec6c741b09"
  },
  {
    "Word Error Rate (WER)": 13.25,
    "code_links": [
      {
        "title": "tensorflow/models",
        "url": "https://github.com/tensorflow/models/tree/master/research/deep_speech"
      },
      {
        "title": "PaddlePaddle/models",
        "url": "https://github.com/PaddlePaddle/models"
      },
      {
        "title": "baidu-research/warp-ctc",
        "url": "https://github.com/baidu-research/warp-ctc"
      },
      {
        "title": "SeanNaren/deepspeech.pytorch",
        "url": "https://github.com/SeanNaren/deepspeech.pytorch"
      },
      {
        "title": "DeepMark/deepmark",
        "url": "https://github.com/DeepMark/deepmark"
      },
      {
        "title": "SeanNaren/deepspeech.torch",
        "url": "https://github.com/SeanNaren/deepspeech.torch"
      },
      {
        "title": "robmsmt/KerasDeepSpeech",
        "url": "https://github.com/robmsmt/KerasDeepSpeech"
      },
      {
        "title": "huylenguyen806/vnasr",
        "url": "https://github.com/huylenguyen806/vnasr"
      },
      {
        "title": "usimarit/TiramisuASR",
        "url": "https://github.com/usimarit/TiramisuASR"
      },
      {
        "title": "usimarit/vnasr",
        "url": "https://github.com/usimarit/vnasr"
      },
      {
        "title": "myrtleSoftware/deepspeech",
        "url": "https://github.com/myrtleSoftware/deepspeech"
      },
      {
        "title": "noahchalifour/deepspeech2-tensorflow",
        "url": "https://github.com/noahchalifour/deepspeech2-tensorflow"
      },
      {
        "title": "raotnameh/End-to-end-E2E-Named-Entity-Recognition-from-English-Speech",
        "url": "https://github.com/raotnameh/End-to-end-E2E-Named-Entity-Recognition-from-English-Speech"
      },
      {
        "title": "UnofficialJuliaMirrorSnapshots/DeepMark-deepmark",
        "url": "https://github.com/UnofficialJuliaMirrorSnapshots/DeepMark-deepmark"
      },
      {
        "title": "UnofficialJuliaMirror/DeepMark-deepmark",
        "url": "https://github.com/UnofficialJuliaMirror/DeepMark-deepmark"
      }
    ],
    "date": "2015-12-08",
    "date2": 20151208,
    "model": "Deep Speech 2",
    "paper": {
      "title": "Deep Speech 2: End-to-End Speech Recognition in English and Mandarin",
      "url": "https://cknow.io/lib/ae2bdef963023028"
    },
    "paper_data_uoa": "ae2bdef963023028"
  },
  {
    "Word Error Rate (WER)": 16.5,
    "code_links": [
      {
        "title": "snipsco/snips-nlu",
        "url": "https://github.com/snipsco/snips-nlu"
      },
      {
        "title": "snipsco/nlu-benchmark",
        "url": "https://github.com/snipsco/nlu-benchmark"
      },
      {
        "title": "MiuLab/SlotGated-SLU",
        "url": "https://github.com/MiuLab/SlotGated-SLU"
      },
      {
        "title": "zliucr/coach",
        "url": "https://github.com/zliucr/coach"
      },
      {
        "title": "valerielimyh/Intent_Recognition_using_BERT",
        "url": "https://github.com/valerielimyh/Intent_Recognition_using_BERT"
      }
    ],
    "date": "2018-05-25",
    "date2": 20180525,
    "model": "Snips",
    "paper": {
      "title": "Snips Voice Platform: an embedded Spoken Language Understanding system for private-by-design voice interfaces",
      "url": "https://cknow.io/lib/4466ad0dd77ea158"
    },
    "paper_data_uoa": "4466ad0dd77ea158"
  }
]