echogarden 0.0.1 → 0.1.0
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- package/README.md +59 -1
- package/data/lexicons/heteronyms.json +674 -0
- package/data/schemas/options.json +1057 -0
- package/data/tables/lcid-table.json +7427 -0
- package/dist/alignment/DTWMfccSequenceAlignment.d.ts +2 -0
- package/dist/alignment/DTWMfccSequenceAlignment.js +26 -0
- package/dist/alignment/DTWMfccSequenceAlignment.js.map +1 -0
- package/dist/alignment/DTWSequenceAlignment.d.ts +5 -0
- package/dist/alignment/DTWSequenceAlignment.js +99 -0
- package/dist/alignment/DTWSequenceAlignment.js.map +1 -0
- package/dist/alignment/DTWSequenceAlignmentWindowed.d.ts +5 -0
- package/dist/alignment/DTWSequenceAlignmentWindowed.js +161 -0
- package/dist/alignment/DTWSequenceAlignmentWindowed.js.map +1 -0
- package/dist/alignment/LevenshteinSequenceAlignment.d.ts +5 -0
- package/dist/alignment/LevenshteinSequenceAlignment.js +102 -0
- package/dist/alignment/LevenshteinSequenceAlignment.js.map +1 -0
- package/dist/alignment/SpeechAlignment.d.ts +27 -0
- package/dist/alignment/SpeechAlignment.js +266 -0
- package/dist/alignment/SpeechAlignment.js.map +1 -0
- package/dist/api/API.d.ts +8 -0
- package/dist/api/API.js +10 -0
- package/dist/api/API.js.map +1 -0
- package/dist/api/APIOptions.d.ts +12 -0
- package/dist/api/APIOptions.js +2 -0
- package/dist/api/APIOptions.js.map +1 -0
- package/dist/api/Alignment.d.ts +32 -0
- package/dist/api/Alignment.js +134 -0
- package/dist/api/Alignment.js.map +1 -0
- package/dist/api/Denoising.d.ts +14 -0
- package/dist/api/Denoising.js +66 -0
- package/dist/api/Denoising.js.map +1 -0
- package/dist/api/Globals.d.ts +1 -0
- package/dist/api/Globals.js +2 -0
- package/dist/api/Globals.js.map +1 -0
- package/dist/api/LanguageDetection.d.ts +49 -0
- package/dist/api/LanguageDetection.js +118 -0
- package/dist/api/LanguageDetection.js.map +1 -0
- package/dist/api/Recognition.d.ts +49 -0
- package/dist/api/Recognition.js +162 -0
- package/dist/api/Recognition.js.map +1 -0
- package/dist/api/Synthesis.d.ts +125 -0
- package/dist/api/Synthesis.js +892 -0
- package/dist/api/Synthesis.js.map +1 -0
- package/dist/api/Translation.d.ts +25 -0
- package/dist/api/Translation.js +69 -0
- package/dist/api/Translation.js.map +1 -0
- package/dist/api/Vad.d.ts +25 -0
- package/dist/api/Vad.js +88 -0
- package/dist/api/Vad.js.map +1 -0
- package/dist/audio/AudioBufferConversion.d.ts +15 -0
- package/dist/audio/AudioBufferConversion.js +228 -0
- package/dist/audio/AudioBufferConversion.js.map +1 -0
- package/dist/audio/AudioPlayer.d.ts +9 -0
- package/dist/audio/AudioPlayer.js +232 -0
- package/dist/audio/AudioPlayer.js.map +1 -0
- package/dist/audio/AudioRecorder.d.ts +3 -0
- package/dist/audio/AudioRecorder.js +68 -0
- package/dist/audio/AudioRecorder.js.map +1 -0
- package/dist/audio/AudioUtilities.d.ts +48 -0
- package/dist/audio/AudioUtilities.js +209 -0
- package/dist/audio/AudioUtilities.js.map +1 -0
- package/dist/audio/SoxPath.d.ts +1 -0
- package/dist/audio/SoxPath.js +19 -0
- package/dist/audio/SoxPath.js.map +1 -0
- package/dist/cli/CLI.d.ts +8 -0
- package/dist/cli/CLI.js +860 -0
- package/dist/cli/CLI.js.map +1 -0
- package/dist/cli/CLIConfigFile.d.ts +3 -0
- package/dist/cli/CLIConfigFile.js +62 -0
- package/dist/cli/CLIConfigFile.js.map +1 -0
- package/dist/cli/CLILauncher.d.ts +2 -0
- package/dist/cli/CLILauncher.js +22 -0
- package/dist/cli/CLILauncher.js.map +1 -0
- package/dist/cli/CLIOptionsSchema.d.ts +5 -0
- package/dist/cli/CLIOptionsSchema.js +36 -0
- package/dist/cli/CLIOptionsSchema.js.map +1 -0
- package/dist/cli/CLIParser.d.ts +6 -0
- package/dist/cli/CLIParser.js +34 -0
- package/dist/cli/CLIParser.js.map +1 -0
- package/dist/cli/CLIStarter.d.ts +1 -0
- package/dist/cli/CLIStarter.js +3 -0
- package/dist/cli/CLIStarter.js.map +1 -0
- package/dist/codecs/FFMpegTranscoder.d.ts +20 -0
- package/dist/codecs/FFMpegTranscoder.js +169 -0
- package/dist/codecs/FFMpegTranscoder.js.map +1 -0
- package/dist/codecs/TIMITCodec.d.ts +9 -0
- package/dist/codecs/TIMITCodec.js +14 -0
- package/dist/codecs/TIMITCodec.js.map +1 -0
- package/dist/codecs/WaveCodec.d.ts +19 -0
- package/dist/codecs/WaveCodec.js +208 -0
- package/dist/codecs/WaveCodec.js.map +1 -0
- package/dist/denoising/RNNoise.d.ts +6 -0
- package/dist/denoising/RNNoise.js +68 -0
- package/dist/denoising/RNNoise.js.map +1 -0
- package/dist/dsp/BiquadFilter.d.ts +26 -0
- package/dist/dsp/BiquadFilter.js +399 -0
- package/dist/dsp/BiquadFilter.js.map +1 -0
- package/dist/dsp/FFT.d.ts +9 -0
- package/dist/dsp/FFT.js +135 -0
- package/dist/dsp/FFT.js.map +1 -0
- package/dist/dsp/MFCC.d.ts +25 -0
- package/dist/dsp/MFCC.js +162 -0
- package/dist/dsp/MFCC.js.map +1 -0
- package/dist/dsp/MelSpectogram.d.ts +19 -0
- package/dist/dsp/MelSpectogram.js +102 -0
- package/dist/dsp/MelSpectogram.js.map +1 -0
- package/dist/dsp/Rubberband.d.ts +52 -0
- package/dist/dsp/Rubberband.js +186 -0
- package/dist/dsp/Rubberband.js.map +1 -0
- package/dist/dsp/Sonic.d.ts +2 -0
- package/dist/dsp/Sonic.js +39 -0
- package/dist/dsp/Sonic.js.map +1 -0
- package/dist/dsp/SpeexResampler.d.ts +3 -0
- package/dist/dsp/SpeexResampler.js +53 -0
- package/dist/dsp/SpeexResampler.js.map +1 -0
- package/dist/math/VectorMath.d.ts +70 -0
- package/dist/math/VectorMath.js +564 -0
- package/dist/math/VectorMath.js.map +1 -0
- package/dist/nlp/ChineseSegmentation.d.ts +1 -0
- package/dist/nlp/ChineseSegmentation.js +53 -0
- package/dist/nlp/ChineseSegmentation.js.map +1 -0
- package/dist/nlp/CompromiseNLP.d.ts +15 -0
- package/dist/nlp/CompromiseNLP.js +66 -0
- package/dist/nlp/CompromiseNLP.js.map +1 -0
- package/dist/nlp/EspeakPhonemizer.d.ts +4 -0
- package/dist/nlp/EspeakPhonemizer.js +133 -0
- package/dist/nlp/EspeakPhonemizer.js.map +1 -0
- package/dist/nlp/IPA.d.ts +19 -0
- package/dist/nlp/IPA.js +113 -0
- package/dist/nlp/IPA.js.map +1 -0
- package/dist/nlp/JapaneseSegmentation.d.ts +1 -0
- package/dist/nlp/JapaneseSegmentation.js +40 -0
- package/dist/nlp/JapaneseSegmentation.js.map +1 -0
- package/dist/nlp/Lexicon.d.ts +17 -0
- package/dist/nlp/Lexicon.js +6 -0
- package/dist/nlp/Lexicon.js.map +1 -0
- package/dist/nlp/PhoneConversion.d.ts +6 -0
- package/dist/nlp/PhoneConversion.js +467 -0
- package/dist/nlp/PhoneConversion.js.map +1 -0
- package/dist/nlp/Segmentation.d.ts +41 -0
- package/dist/nlp/Segmentation.js +158 -0
- package/dist/nlp/Segmentation.js.map +1 -0
- package/dist/nlp/TextNormalizer.d.ts +4 -0
- package/dist/nlp/TextNormalizer.js +69 -0
- package/dist/nlp/TextNormalizer.js.map +1 -0
- package/dist/recognition/AmazonTranscribeSTT.d.ts +6 -0
- package/dist/recognition/AmazonTranscribeSTT.js +79 -0
- package/dist/recognition/AmazonTranscribeSTT.js.map +1 -0
- package/dist/recognition/AzureCognitiveServicesSTT.d.ts +7 -0
- package/dist/recognition/AzureCognitiveServicesSTT.js +51 -0
- package/dist/recognition/AzureCognitiveServicesSTT.js.map +1 -0
- package/dist/recognition/GoogleCloudSTT.d.ts +7 -0
- package/dist/recognition/GoogleCloudSTT.js +66 -0
- package/dist/recognition/GoogleCloudSTT.js.map +1 -0
- package/dist/recognition/SileroSTT.d.ts +9 -0
- package/dist/recognition/SileroSTT.js +125 -0
- package/dist/recognition/SileroSTT.js.map +1 -0
- package/dist/recognition/VoskSTT.d.ts +10 -0
- package/dist/recognition/VoskSTT.js +78 -0
- package/dist/recognition/VoskSTT.js.map +1 -0
- package/dist/recognition/WhisperSTT.d.ts +69 -0
- package/dist/recognition/WhisperSTT.js +977 -0
- package/dist/recognition/WhisperSTT.js.map +1 -0
- package/dist/server/Server.d.ts +1 -0
- package/dist/server/Server.js +18 -0
- package/dist/server/Server.js.map +1 -0
- package/dist/speech-language-detection/SileroLanguageDetection.d.ts +9 -0
- package/dist/speech-language-detection/SileroLanguageDetection.js +47 -0
- package/dist/speech-language-detection/SileroLanguageDetection.js.map +1 -0
- package/dist/subtitles/Subtitles.d.ts +25 -0
- package/dist/subtitles/Subtitles.js +286 -0
- package/dist/subtitles/Subtitles.js.map +1 -0
- package/dist/synthesis/AwsPollyTTS.d.ts +7 -0
- package/dist/synthesis/AwsPollyTTS.js +51 -0
- package/dist/synthesis/AwsPollyTTS.js.map +1 -0
- package/dist/synthesis/AzureCognitiveServicesTTS.d.ts +9 -0
- package/dist/synthesis/AzureCognitiveServicesTTS.js +103 -0
- package/dist/synthesis/AzureCognitiveServicesTTS.js.map +1 -0
- package/dist/synthesis/CoquiServerTTS.d.ts +6 -0
- package/dist/synthesis/CoquiServerTTS.js +22 -0
- package/dist/synthesis/CoquiServerTTS.js.map +1 -0
- package/dist/synthesis/ElevenLabsTTS.d.ts +8 -0
- package/dist/synthesis/ElevenLabsTTS.js +48 -0
- package/dist/synthesis/ElevenLabsTTS.js.map +1 -0
- package/dist/synthesis/EspeakTTS.d.ts +46 -0
- package/dist/synthesis/EspeakTTS.js +353 -0
- package/dist/synthesis/EspeakTTS.js.map +1 -0
- package/dist/synthesis/FliteTTS.d.ts +17 -0
- package/dist/synthesis/FliteTTS.js +326 -0
- package/dist/synthesis/FliteTTS.js.map +1 -0
- package/dist/synthesis/GoogleCloudTTS.d.ts +16 -0
- package/dist/synthesis/GoogleCloudTTS.js +72 -0
- package/dist/synthesis/GoogleCloudTTS.js.map +1 -0
- package/dist/synthesis/GoogleTranslateTTS.d.ts +13 -0
- package/dist/synthesis/GoogleTranslateTTS.js +177 -0
- package/dist/synthesis/GoogleTranslateTTS.js.map +1 -0
- package/dist/synthesis/MicrosoftEdgeTTS.d.ts +10 -0
- package/dist/synthesis/MicrosoftEdgeTTS.js +216 -0
- package/dist/synthesis/MicrosoftEdgeTTS.js.map +1 -0
- package/dist/synthesis/SamTTS.d.ts +4 -0
- package/dist/synthesis/SamTTS.js +21 -0
- package/dist/synthesis/SamTTS.js.map +1 -0
- package/dist/synthesis/SapiTTS.d.ts +9 -0
- package/dist/synthesis/SapiTTS.js +211 -0
- package/dist/synthesis/SapiTTS.js.map +1 -0
- package/dist/synthesis/StreamlabsPollyTTS.d.ts +12 -0
- package/dist/synthesis/StreamlabsPollyTTS.js +88 -0
- package/dist/synthesis/StreamlabsPollyTTS.js.map +1 -0
- package/dist/synthesis/SvoxPicoTTS.d.ts +11 -0
- package/dist/synthesis/SvoxPicoTTS.js +236 -0
- package/dist/synthesis/SvoxPicoTTS.js.map +1 -0
- package/dist/synthesis/VitsTTS.d.ts +27 -0
- package/dist/synthesis/VitsTTS.js +359 -0
- package/dist/synthesis/VitsTTS.js.map +1 -0
- package/dist/tests/Test.d.ts +1 -0
- package/dist/tests/Test.js +10 -0
- package/dist/tests/Test.js.map +1 -0
- package/dist/text-language-detection/FastTextLanguageDetection.d.ts +2 -0
- package/dist/text-language-detection/FastTextLanguageDetection.js +38 -0
- package/dist/text-language-detection/FastTextLanguageDetection.js.map +1 -0
- package/dist/text-language-detection/TinyLDLanguageDetection.d.ts +2 -0
- package/dist/text-language-detection/TinyLDLanguageDetection.js +12 -0
- package/dist/text-language-detection/TinyLDLanguageDetection.js.map +1 -0
- package/dist/utilities/BinaryArrayConversion.d.ts +15 -0
- package/dist/utilities/BinaryArrayConversion.js +115 -0
- package/dist/utilities/BinaryArrayConversion.js.map +1 -0
- package/dist/utilities/Compression.d.ts +4 -0
- package/dist/utilities/Compression.js +67 -0
- package/dist/utilities/Compression.js.map +1 -0
- package/dist/utilities/FileDownloader.d.ts +3 -0
- package/dist/utilities/FileDownloader.js +147 -0
- package/dist/utilities/FileDownloader.js.map +1 -0
- package/dist/utilities/FileSystem.d.ts +30 -0
- package/dist/utilities/FileSystem.js +141 -0
- package/dist/utilities/FileSystem.js.map +1 -0
- package/dist/utilities/Hashing.d.ts +10 -0
- package/dist/utilities/Hashing.js +169 -0
- package/dist/utilities/Hashing.js.map +1 -0
- package/dist/utilities/Locale.d.ts +9 -0
- package/dist/utilities/Locale.js +65 -0
- package/dist/utilities/Locale.js.map +1 -0
- package/dist/utilities/Logger.d.ts +11 -0
- package/dist/utilities/Logger.js +49 -0
- package/dist/utilities/Logger.js.map +1 -0
- package/dist/utilities/NdArrayUtilities.d.ts +3 -0
- package/dist/utilities/NdArrayUtilities.js +22 -0
- package/dist/utilities/NdArrayUtilities.js.map +1 -0
- package/dist/utilities/ObjectUtilities.d.ts +4 -0
- package/dist/utilities/ObjectUtilities.js +132 -0
- package/dist/utilities/ObjectUtilities.js.map +1 -0
- package/dist/utilities/OpenPromise.d.ts +6 -0
- package/dist/utilities/OpenPromise.js +12 -0
- package/dist/utilities/OpenPromise.js.map +1 -0
- package/dist/utilities/PackageManager.d.ts +4 -0
- package/dist/utilities/PackageManager.js +46 -0
- package/dist/utilities/PackageManager.js.map +1 -0
- package/dist/utilities/RandomGenerator.d.ts +35 -0
- package/dist/utilities/RandomGenerator.js +149 -0
- package/dist/utilities/RandomGenerator.js.map +1 -0
- package/dist/utilities/TarballMaker.d.ts +4 -0
- package/dist/utilities/TarballMaker.js +50 -0
- package/dist/utilities/TarballMaker.js.map +1 -0
- package/dist/utilities/Timeline.d.ts +20 -0
- package/dist/utilities/Timeline.js +110 -0
- package/dist/utilities/Timeline.js.map +1 -0
- package/dist/utilities/Timer.d.ts +13 -0
- package/dist/utilities/Timer.js +70 -0
- package/dist/utilities/Timer.js.map +1 -0
- package/dist/utilities/Utilities.d.ts +68 -0
- package/dist/utilities/Utilities.js +305 -0
- package/dist/utilities/Utilities.js.map +1 -0
- package/dist/utilities/WasmMemoryManager.d.ts +142 -0
- package/dist/utilities/WasmMemoryManager.js +407 -0
- package/dist/utilities/WasmMemoryManager.js.map +1 -0
- package/dist/utilities/WebReader.d.ts +1 -0
- package/dist/utilities/WebReader.js +47 -0
- package/dist/utilities/WebReader.js.map +1 -0
- package/dist/utilities/WikipediaReader.d.ts +1 -0
- package/dist/utilities/WikipediaReader.js +31 -0
- package/dist/utilities/WikipediaReader.js.map +1 -0
- package/dist/voice-activity-detection/SileroVAD.d.ts +13 -0
- package/dist/voice-activity-detection/SileroVAD.js +58 -0
- package/dist/voice-activity-detection/SileroVAD.js.map +1 -0
- package/dist/voice-activity-detection/WebRtcVAD.d.ts +3 -0
- package/dist/voice-activity-detection/WebRtcVAD.js +53 -0
- package/dist/voice-activity-detection/WebRtcVAD.js.map +1 -0
- package/docs/CLI.md +234 -0
- package/docs/Development.md +3 -0
- package/docs/Engines.md +80 -0
- package/docs/Licenses.md +37 -0
- package/docs/Options.md +192 -0
- package/docs/Roadmap.md +16 -0
- package/docs/Technical.md +72 -0
- package/package.json +119 -19
- package/cli.js +0 -3
|
@@ -0,0 +1 @@
|
|
|
1
|
+
{"version":3,"file":"Segmentation.js","sourceRoot":"","sources":["../../src/nlp/Segmentation.ts"],"names":[],"mappings":"AAAA,OAAO,OAAO,MAAM,SAAS,CAAA;AAE7B,OAAO,KAAK,gBAAgB,MAAM,mBAAmB,CAAA;AACrD,OAAO,EAAE,6BAA6B,EAAE,MAAM,0BAA0B,CAAA;AAExE,OAAO,EAAE,QAAQ,EAAE,aAAa,EAAE,YAAY,EAAE,WAAW,EAAE,MAAM,2BAA2B,CAAA;AAC9F,OAAO,EAAE,oBAAoB,EAAE,MAAM,wBAAwB,CAAA;AAC7D,OAAO,EAAE,iCAAiC,EAAE,MAAM,2BAA2B,CAAA;AAE7E,MAAM,GAAG,GAAG,WAAW,CAAA;AAEvB,MAAM,CAAC,MAAM,oBAAoB,GAAG,OAAO,CAAC,0BAA0B,CAAC,CAAA;AACvE,MAAM,CAAC,MAAM,gBAAgB,GAAG,CAAC,GAAG,EAAE,GAAG,EAAE,GAAG,CAAC,CAAA;AAC/C,MAAM,CAAC,MAAM,kBAAkB,GAAG,CAAC,GAAG,EAAE,GAAG,EAAE,GAAG,CAAC,CAAA;AACjD,MAAM,CAAC,MAAM,WAAW,GAAG,CAAC,GAAG,EAAE,GAAG,EAAE,GAAG,EAAE,GAAG,EAAE,GAAG,EAAE,GAAG,EAAE,GAAG,EAAE,GAAG,CAAC,CAAA;AAEnE,MAAM,UAAU,MAAM,CAAC,GAAW;IACjC,GAAG,GAAG,GAAG,CAAC,IAAI,EAAE,CAAA;IAChB,OAAO,oBAAoB,CAAC,IAAI,CAAC,GAAG,CAAC,IAAI,WAAW,CAAC,QAAQ,CAAC,GAAG,CAAC,CAAA;AACnE,CAAC;AAED,MAAM,OAAO,QAAQ;IACpB,OAAO,GAAa,EAAE,CAAA;IAEb,mBAAmB,GAAG,IAAI,CAAA;IAEnC,IAAI,MAAM,KAAK,OAAO,QAAQ,CAAC,IAAI,CAAC,OAAO,EAAE,CAAC,MAAM,EAAE,EAAE,CAAC,MAAM,CAAC,MAAM,CAAC,CAAA,CAAC,CAAC;IAEzE,IAAI,IAAI,KAAK,OAAO,IAAI,CAAC,OAAO,CAAC,MAAM,CAAS,CAAC,MAAM,EAAE,MAAM,EAAE,EAAE,CAAC,MAAM,GAAG,MAAM,CAAC,IAAI,EAAE,EAAE,CAAC,CAAA,CAAC,CAAC;CAC/F;AAED,MAAM,OAAO,MAAM;IAClB,KAAK,GAAW,EAAE,CAAA;IAElB,IAAI,MAAM,KAAK,OAAO,QAAQ,CAAC,IAAI,CAAC,KAAK,EAAE,CAAC,IAAI,EAAE,EAAE,CAAC,IAAI,CAAC,MAAM,CAAC,CAAA,CAAC,CAAC;IAEnE,IAAI,IAAI,KAAK,OAAO,IAAI,CAAC,KAAK,CAAC,MAAM,CAAS,CAAC,MAAM,EAAE,IAAI,EAAE,EAAE,CAAC,MAAM,GAAG,IAAI,CAAC,IAAI,EAAE,EAAE,CAAC,CAAA,CAAC,CAAC;IAEzF,IAAI,QAAQ;QACX,IAAI,IAAI,CAAC,KAAK,CAAC,MAAM,IAAI,CAAC,EAAE;YAC3B,OAAO,SAAS,CAAA;SAChB;QAED,OAAO,IAAI,CAAC,KAAK,CAAC,IAAI,CAAC,KAAK,CAAC,MAAM,GAAG,CAAC,CAAC,CAAA;IACzC,CAAC;IAED,IAAI,mBAAmB,KAAK,OAAO,IAAI,CAAC,QAAQ,IAAI,IAAI,CAAC,CAAC,CAAC,IAAI,CAAC,QAAQ,CAAC,mBAAmB,CAAC,CAAC,CAAC,KAAK,CAAA,CAAC,CAAC;CACtG;AAED,MAAM,OAAO,IAAI;IACP,IAAI,CAAQ;IACrB,mBAAmB,CAAS;IAE5B,YAAY,IAAY,EAAE,mBAA4B;QACrD,IAAI,CAAC,IAAI,GAAG,IAAI,CAAA;QAChB,IAAI,CAAC,mBAAmB,GAAG,mBAAmB,CAAA;IAC/C,CAAC;IAED,IAAI,uBAAuB,KAAK,OAAO,CAAC,oBAAoB,CAAC,IAAI,CAAC,IAAI,CAAC,IAAI,CAAC,IAAI,CAAC,IAAI,CAAC,YAAY,CAAA,CAAC,CAAC;IAEpG,IAAI,YAAY,KAAK,OAAO,WAAW,CAAC,QAAQ,CAAC,IAAI,CAAC,IAAI,CAAC,CAAA,CAAC,CAAC;IAE7D,IAAI,iBAAiB,KAAK,OAAO,IAAI,CAAC,uBAAuB,IAAI,aAAa,CAAC,IAAI,CAAC,IAAI,EAAE,gBAAgB,CAAC,CAAA,CAAC,CAAC;IAE7G,IAAI,MAAM,KAAK,OAAO,IAAI,CAAC,IAAI,CAAC,MAAM,CAAA,CAAC,CAAC;CACxC;AAID,MAAM,OAAO,QAAQ;IACpB,QAAQ,GAAc,EAAE,CAAA;IAExB,IAAI,MAAM,KAAK,OAAO,QAAQ,CAAC,IAAI,CAAC,QAAQ,EAAE,CAAC,MAAM,EAAE,EAAE,CAAC,MAAM,CAAC,MAAM,CAAC,CAAA,CAAC,CAAC;IAE1E,IAAI,IAAI,KAAK,OAAO,IAAI,CAAC,QAAQ,CAAC,MAAM,CAAS,CAAC,MAAM,EAAE,OAAO,EAAE,EAAE,CAAC,MAAM,GAAG,OAAO,CAAC,IAAI,EAAE,EAAE,CAAC,CAAA,CAAC,CAAC;IAElG,IAAI,OAAO,KAAK,OAAO,IAAI,CAAC,MAAM,IAAI,CAAC,CAAA,CAAC,CAAC;IAEzC,IAAI,UAAU,KAAK,OAAO,CAAC,IAAI,CAAC,OAAO,CAAA,CAAC,CAAC;IAEzC,IAAI,WAAW;QACd,IAAI,IAAI,CAAC,OAAO,EAAE;YACjB,OAAO,SAAS,CAAA;SAChB;QAED,OAAO,IAAI,CAAC,QAAQ,CAAC,IAAI,CAAC,QAAQ,CAAC,MAAM,GAAG,CAAC,CAAC,CAAA;IAC/C,CAAC;CACD;AAED,MAAM,CAAC,KAAK,UAAU,gBAAgB,CAAC,IAAY,EAAE,iBAAyB,EAAE,QAAgB,EAAE,iBAAiB,GAAG,IAAI,EAAE,eAAe,GAAG,IAAI;IACjJ,MAAM,UAAU,GAAG,MAAM,KAAK,CAAC,IAAI,EAAE,QAAQ,CAAC,CAAA;IAE9C,MAAM,SAAS,GAAe,EAAE,CAAA;IAChC,IAAI,eAAe,GAAG,IAAI,QAAQ,EAAE,CAAA;IAEpC,MAAM,oCAAoC,GAAG,GAAG,EAAE,CAAC,iBAAiB,GAAG,eAAe,CAAC,MAAM,CAAA;IAC7F,MAAM,yBAAyB,GAAG,GAAG,EAAE;QACtC,IAAI,eAAe,CAAC,UAAU,EAAE;YAC/B,SAAS,CAAC,IAAI,CAAC,eAAe,CAAC,CAAA;YAC/B,eAAe,GAAG,IAAI,QAAQ,EAAE,CAAA;SAChC;IACF,CAAC,CAAA;IAED,MAAM,mBAAmB,GAAG,CAAC,OAAgB,EAAE,EAAE,CAAC,OAAO,CAAC,MAAM,IAAI,oCAAoC,EAAE,CAAA;IAE1G,KAAK,MAAM,QAAQ,IAAI,UAAU,EAAE;QAClC,IAAI,mBAAmB,CAAC,QAAQ,CAAC,EAAE;YAClC,eAAe,CAAC,QAAQ,CAAC,IAAI,CAAC,QAAQ,CAAC,CAAA;YACvC,SAAQ;SACR;QAED,IAAI,iBAAiB,EAAE;YACtB,yBAAyB,EAAE,CAAA;YAE3B,IAAI,mBAAmB,CAAC,QAAQ,CAAC,EAAE;gBAClC,eAAe,CAAC,QAAQ,CAAC,IAAI,CAAC,QAAQ,CAAC,CAAA;gBACvC,SAAQ;aACR;SACD;QAED,KAAK,MAAM,MAAM,IAAI,QAAQ,CAAC,OAAO,EAAE;YACtC,IAAI,mBAAmB,CAAC,MAAM,CAAC,EAAE;gBAChC,eAAe,CAAC,QAAQ,CAAC,IAAI,CAAC,MAAM,CAAC,CAAA;gBACrC,SAAQ;aACR;YAGD,IAAI,eAAe,EAAE;gBACpB,yBAAyB,EAAE,CAAA;gBAE3B,IAAI,mBAAmB,CAAC,MAAM,CAAC,EAAE;oBAChC,eAAe,CAAC,QAAQ,CAAC,IAAI,CAAC,MAAM,CAAC,CAAA;oBACrC,SAAQ;iBACR;aACD;YAED,KAAK,MAAM,IAAI,IAAI,MAAM,CAAC,KAAK,EAAE;gBAChC,IAAI,mBAAmB,CAAC,IAAI,CAAC,EAAE;oBAC9B,eAAe,CAAC,QAAQ,CAAC,IAAI,CAAC,IAAI,CAAC,CAAA;oBACnC,SAAQ;iBACR;gBAED,yBAAyB,EAAE,CAAA;gBAE3B,IAAI,mBAAmB,CAAC,IAAI,CAAC,EAAE;oBAC9B,eAAe,CAAC,QAAQ,CAAC,IAAI,CAAC,IAAI,CAAC,CAAA;oBACnC,SAAQ;iBACR;gBAED,MAAM,IAAI,KAAK,CAAC,gCAAgC,IAAI,CAAC,MAAM,mDAAmD,iBAAiB,EAAE,CAAC,CAAA;aAClI;SACD;KACD;IAED,yBAAyB,EAAE,CAAA;IAE3B,OAAO,SAAS,CAAA;AACjB,CAAC;AAED,MAAM,CAAC,KAAK,UAAU,KAAK,CAAC,IAAY,EAAE,QAAgB;IACzD,MAAM,aAAa,GAAG,gBAAgB,CAAC,IAAI,EAAE,QAAQ,CAAC,CAAA;IACtD,MAAM,SAAS,GAAe,EAAE,CAAA;IAEhC,KAAK,MAAM,YAAY,IAAI,aAAa,EAAE;QACzC,MAAM,QAAQ,GAAG,IAAI,QAAQ,EAAE,CAAA;QAE/B,IAAI,aAAa,GAAG,IAAI,MAAM,EAAE,CAAA;QAChC,MAAM,SAAS,GAAG,MAAM,YAAY,CAAC,YAAY,EAAE,QAAQ,CAAC,CAAA;QAE5D,KAAK,IAAI,SAAS,GAAG,CAAC,EAAE,SAAS,GAAG,SAAS,CAAC,MAAM,EAAE,SAAS,EAAE,EAAE;YAClE,MAAM,IAAI,GAAG,IAAI,IAAI,CAAC,SAAS,CAAC,SAAS,CAAC,EAAE,SAAS,IAAI,SAAS,CAAC,MAAM,GAAG,CAAC,CAAC,CAAA;YAE9E,IAAI,IAAI,CAAC,iBAAiB,EAAE;gBAC3B,MAAM,cAAc,GAAG,YAAY,CAAC,IAAI,CAAC,IAAI,EAAE,gBAAgB,CAAC,CAAA;gBAChE,aAAa,CAAC,KAAK,CAAC,IAAI,CAAC,IAAI,IAAI,CAAC,IAAI,CAAC,IAAI,CAAC,SAAS,CAAC,CAAC,EAAE,cAAc,GAAG,CAAC,CAAC,EAAE,IAAI,CAAC,mBAAmB,CAAC,CAAC,CAAA;gBACxG,QAAQ,CAAC,OAAO,CAAC,IAAI,CAAC,aAAa,CAAC,CAAA;gBAEpC,aAAa,GAAG,IAAI,MAAM,EAAE,CAAA;gBAC5B,aAAa,CAAC,KAAK,CAAC,IAAI,CAAC,IAAI,IAAI,CAAC,IAAI,CAAC,IAAI,CAAC,SAAS,CAAC,cAAc,GAAG,CAAC,CAAC,EAAE,KAAK,CAAC,CAAC,CAAA;aAClF;iBAAM;gBACN,aAAa,CAAC,KAAK,CAAC,IAAI,CAAC,IAAI,CAAC,CAAA;aAC9B;SACD;QAED,IAAI,aAAa,CAAC,KAAK,CAAC,MAAM,GAAG,CAAC,EAAE;YACnC,QAAQ,CAAC,OAAO,CAAC,IAAI,CAAC,aAAa,CAAC,CAAA;SACpC;QAED,SAAS,CAAC,IAAI,CAAC,QAAQ,CAAC,CAAA;KACxB;IAED,OAAO,SAAS,CAAA;AACjB,CAAC;AAED,MAAM,UAAU,gBAAgB,CAAC,IAAY,EAAE,QAAgB;IAC9D,MAAM,aAAa,GAAG,oBAAoB,CAAC,QAAQ,IAAI,EAAE,CAAC,CAAA;IAE1D,OAAO,gBAAgB,CAAC,aAAa,CAAC,IAAI,EAAE,gBAAgB,CAAC,YAAY,CAAC,aAAa,CAAC,CAAC,CAAA;AAC1F,CAAC;AAGD,MAAM,CAAC,KAAK,UAAU,YAAY,CAAC,IAAY,EAAE,QAAgB;IAChE,MAAM,aAAa,GAAG,oBAAoB,CAAC,QAAQ,IAAI,EAAE,CAAC,CAAA;IAE1D,IAAI,aAAa,IAAI,IAAI,IAAI,aAAa,IAAI,KAAK,EAAE;QACpD,OAAO,6BAA6B,CAAC,IAAI,EAAE,IAAI,EAAE,IAAI,CAAC,CAAA;KACtD;SAAM,IAAI,aAAa,IAAI,IAAI,EAAE;QACjC,OAAO,iCAAiC,CAAC,IAAI,CAAC,CAAA;KAC9C;SAAM;QACN,OAAO,gBAAgB,CAAC,SAAS,CAAC,IAAI,EAAE,gBAAgB,CAAC,YAAY,CAAC,aAAa,CAAC,CAAC,CAAA;KACrF;AACF,CAAC;AAED,MAAM,UAAU,iBAAiB,CAAC,IAAY;IAC7C,OAAO,IAAI,CAAC,KAAK,CAAC,WAAW,CAAC,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,CAAC,CAAC,CAAC,IAAI,EAAE,CAAC,CAAC,MAAM,CAAC,CAAC,CAAC,EAAE,CAAC,CAAC,CAAC,MAAM,GAAG,CAAC,CAAC,CAAA;AAC5E,CAAC"}
|
|
@@ -0,0 +1,4 @@
|
|
|
1
|
+
import { CompromiseParsedSentence } from "./CompromiseNLP.js";
|
|
2
|
+
export declare function getNormalizationMapForSpeech(terms: CompromiseParsedSentence, language: string): Map<number, string>;
|
|
3
|
+
export declare function normalizeFourDigitYearString(yearString: string): string;
|
|
4
|
+
export declare function normalizeFourDigitDecadeString(decadeString: string): string;
|
|
@@ -0,0 +1,69 @@
|
|
|
1
|
+
import { getShortLanguageCode } from "../utilities/Locale.js";
|
|
2
|
+
export function getNormalizationMapForSpeech(terms, language) {
|
|
3
|
+
language = getShortLanguageCode(language);
|
|
4
|
+
const normalizationMap = new Map();
|
|
5
|
+
if (language != "en") {
|
|
6
|
+
return normalizationMap;
|
|
7
|
+
}
|
|
8
|
+
const numberPattern = /^[0-9][0-9\,\.]*$/;
|
|
9
|
+
const fourDigitYearPattern = /^[0-9][0-9][0-9][0-9]$/;
|
|
10
|
+
const fourDigitDecadePattern = /^[0-9][0-9][0-9]0s$/;
|
|
11
|
+
const fourDigitYearRangePattern = /^[0-9][0-9][0-9][0-9][\-\–][0-9][0-9][0-9][0-9]$/;
|
|
12
|
+
for (let termIndex = 0; termIndex < terms.length; termIndex++) {
|
|
13
|
+
const term = terms[termIndex];
|
|
14
|
+
const termText = term.text;
|
|
15
|
+
const lowerCaseTermText = termText.toLocaleLowerCase();
|
|
16
|
+
const nextTerms = terms.slice(termIndex + 1);
|
|
17
|
+
const nextTerm = nextTerms[0];
|
|
18
|
+
const nextTermText = nextTerm?.text;
|
|
19
|
+
if (lowerCaseTermText == "in" &&
|
|
20
|
+
term.postText.length > 0 && term.postText.trim() == "" &&
|
|
21
|
+
fourDigitYearPattern.test(nextTermText)) {
|
|
22
|
+
const normalizedString = normalizeFourDigitYearString(nextTermText);
|
|
23
|
+
normalizationMap.set(termIndex + 1, normalizedString);
|
|
24
|
+
termIndex += 1;
|
|
25
|
+
}
|
|
26
|
+
else if (lowerCaseTermText == "the" &&
|
|
27
|
+
term.postText.length > 0 && term.postText.trim() == "" &&
|
|
28
|
+
fourDigitDecadePattern.test(nextTermText)) {
|
|
29
|
+
const normalizedString = normalizeFourDigitDecadeString(nextTermText);
|
|
30
|
+
normalizationMap.set(termIndex + 1, normalizedString);
|
|
31
|
+
termIndex += 1;
|
|
32
|
+
}
|
|
33
|
+
else if (fourDigitYearRangePattern.test(termText)) {
|
|
34
|
+
const startYearString = normalizeFourDigitYearString(termText.substring(0, 4));
|
|
35
|
+
const endYearString = normalizeFourDigitYearString(termText.substring(5, 9));
|
|
36
|
+
const normalizedString = `${startYearString} to ${endYearString}`;
|
|
37
|
+
normalizationMap.set(termIndex, normalizedString);
|
|
38
|
+
}
|
|
39
|
+
}
|
|
40
|
+
return normalizationMap;
|
|
41
|
+
}
|
|
42
|
+
export function normalizeFourDigitYearString(yearString) {
|
|
43
|
+
const firstTwoDigitsValue = parseFloat(yearString.substring(0, 2));
|
|
44
|
+
const secondTwoDigitsValue = parseFloat(yearString.substring(2, 4));
|
|
45
|
+
let normalizedString;
|
|
46
|
+
if (firstTwoDigitsValue >= 10 && secondTwoDigitsValue >= 10) {
|
|
47
|
+
normalizedString = `${firstTwoDigitsValue} ${secondTwoDigitsValue}`;
|
|
48
|
+
}
|
|
49
|
+
else if (firstTwoDigitsValue >= 10 && firstTwoDigitsValue % 10 != 0 && secondTwoDigitsValue < 10) {
|
|
50
|
+
normalizedString = `${firstTwoDigitsValue} oh ${secondTwoDigitsValue}`;
|
|
51
|
+
}
|
|
52
|
+
else {
|
|
53
|
+
normalizedString = yearString;
|
|
54
|
+
}
|
|
55
|
+
return normalizedString;
|
|
56
|
+
}
|
|
57
|
+
export function normalizeFourDigitDecadeString(decadeString) {
|
|
58
|
+
const firstTwoDigitsValue = parseFloat(decadeString.substring(0, 2));
|
|
59
|
+
const secondTwoDigitsValue = parseFloat(decadeString.substring(2, 4));
|
|
60
|
+
let normalizedString;
|
|
61
|
+
if (firstTwoDigitsValue >= 10 && firstTwoDigitsValue % 10 != 0) {
|
|
62
|
+
normalizedString = `${firstTwoDigitsValue} ${secondTwoDigitsValue}s`;
|
|
63
|
+
}
|
|
64
|
+
else {
|
|
65
|
+
normalizedString = decadeString;
|
|
66
|
+
}
|
|
67
|
+
return normalizedString;
|
|
68
|
+
}
|
|
69
|
+
//# sourceMappingURL=TextNormalizer.js.map
|
|
@@ -0,0 +1 @@
|
|
|
1
|
+
{"version":3,"file":"TextNormalizer.js","sourceRoot":"","sources":["../../src/nlp/TextNormalizer.ts"],"names":[],"mappings":"AAAA,OAAO,EAAE,oBAAoB,EAAE,MAAM,wBAAwB,CAAA;AAG7D,MAAM,UAAU,4BAA4B,CAAC,KAA+B,EAAE,QAAgB;IAC7F,QAAQ,GAAG,oBAAoB,CAAC,QAAQ,CAAC,CAAA;IAEzC,MAAM,gBAAgB,GAAG,IAAI,GAAG,EAAkB,CAAA;IAElD,IAAI,QAAQ,IAAI,IAAI,EAAE;QACrB,OAAO,gBAAgB,CAAA;KACvB;IAED,MAAM,aAAa,GAAG,mBAAmB,CAAA;IAEzC,MAAM,oBAAoB,GAAG,wBAAwB,CAAA;IACrD,MAAM,sBAAsB,GAAG,qBAAqB,CAAA;IAEpD,MAAM,yBAAyB,GAAG,kDAAkD,CAAA;IAEpF,KAAK,IAAI,SAAS,GAAG,CAAC,EAAE,SAAS,GAAG,KAAK,CAAC,MAAM,EAAE,SAAS,EAAE,EAAE;QAC9D,MAAM,IAAI,GAAG,KAAK,CAAC,SAAS,CAAC,CAAA;QAC7B,MAAM,QAAQ,GAAG,IAAI,CAAC,IAAI,CAAA;QAC1B,MAAM,iBAAiB,GAAG,QAAQ,CAAC,iBAAiB,EAAE,CAAA;QAEtD,MAAM,SAAS,GAAG,KAAK,CAAC,KAAK,CAAC,SAAS,GAAG,CAAC,CAAC,CAAA;QAC5C,MAAM,QAAQ,GAAG,SAAS,CAAC,CAAC,CAAC,CAAA;QAC7B,MAAM,YAAY,GAAG,QAAQ,EAAE,IAAI,CAAA;QAEnC,IACC,iBAAiB,IAAI,IAAI;YACzB,IAAI,CAAC,QAAQ,CAAC,MAAM,GAAG,CAAC,IAAI,IAAI,CAAC,QAAQ,CAAC,IAAI,EAAE,IAAI,EAAE;YACtD,oBAAoB,CAAC,IAAI,CAAC,YAAY,CAAC,EAAE;YAEzC,MAAM,gBAAgB,GAAG,4BAA4B,CAAC,YAAY,CAAC,CAAA;YAEnE,gBAAgB,CAAC,GAAG,CAAC,SAAS,GAAG,CAAC,EAAE,gBAAgB,CAAC,CAAA;YAErD,SAAS,IAAI,CAAC,CAAA;SACd;aAAM,IACN,iBAAiB,IAAI,KAAK;YAC1B,IAAI,CAAC,QAAQ,CAAC,MAAM,GAAG,CAAC,IAAI,IAAI,CAAC,QAAQ,CAAC,IAAI,EAAE,IAAI,EAAE;YACtD,sBAAsB,CAAC,IAAI,CAAC,YAAY,CAAC,EAAE;YAE3C,MAAM,gBAAgB,GAAG,8BAA8B,CAAC,YAAY,CAAC,CAAA;YAErE,gBAAgB,CAAC,GAAG,CAAC,SAAS,GAAG,CAAC,EAAE,gBAAgB,CAAC,CAAA;YAErD,SAAS,IAAI,CAAC,CAAA;SACd;aAAM,IAAI,yBAAyB,CAAC,IAAI,CAAC,QAAQ,CAAC,EAAE;YACpD,MAAM,eAAe,GAAG,4BAA4B,CAAC,QAAQ,CAAC,SAAS,CAAC,CAAC,EAAE,CAAC,CAAC,CAAC,CAAA;YAC9E,MAAM,aAAa,GAAG,4BAA4B,CAAC,QAAQ,CAAC,SAAS,CAAC,CAAC,EAAE,CAAC,CAAC,CAAC,CAAA;YAE5E,MAAM,gBAAgB,GAAG,GAAG,eAAe,OAAO,aAAa,EAAE,CAAA;YAEjE,gBAAgB,CAAC,GAAG,CAAC,SAAS,EAAE,gBAAgB,CAAC,CAAA;SACjD;KACD;IAED,OAAO,gBAAgB,CAAA;AACxB,CAAC;AAED,MAAM,UAAU,4BAA4B,CAAC,UAAkB;IAC9D,MAAM,mBAAmB,GAAG,UAAU,CAAC,UAAU,CAAC,SAAS,CAAC,CAAC,EAAE,CAAC,CAAC,CAAC,CAAA;IAClE,MAAM,oBAAoB,GAAG,UAAU,CAAC,UAAU,CAAC,SAAS,CAAC,CAAC,EAAE,CAAC,CAAC,CAAC,CAAA;IAEnE,IAAI,gBAAwB,CAAA;IAE5B,IAAI,mBAAmB,IAAI,EAAE,IAAI,oBAAoB,IAAI,EAAE,EAAE;QAC5D,gBAAgB,GAAG,GAAG,mBAAmB,IAAI,oBAAoB,EAAE,CAAA;KACnE;SAAM,IAAI,mBAAmB,IAAI,EAAE,IAAI,mBAAmB,GAAG,EAAE,IAAI,CAAC,IAAI,oBAAoB,GAAG,EAAE,EAAE;QACnG,gBAAgB,GAAG,GAAG,mBAAmB,OAAO,oBAAoB,EAAE,CAAA;KACtE;SAAM;QACN,gBAAgB,GAAG,UAAU,CAAA;KAC7B;IAED,OAAO,gBAAgB,CAAA;AACxB,CAAC;AAED,MAAM,UAAU,8BAA8B,CAAC,YAAoB;IAClE,MAAM,mBAAmB,GAAG,UAAU,CAAC,YAAY,CAAC,SAAS,CAAC,CAAC,EAAE,CAAC,CAAC,CAAC,CAAA;IACpE,MAAM,oBAAoB,GAAG,UAAU,CAAC,YAAY,CAAC,SAAS,CAAC,CAAC,EAAE,CAAC,CAAC,CAAC,CAAA;IAErE,IAAI,gBAAwB,CAAA;IAE5B,IAAI,mBAAmB,IAAI,EAAE,IAAI,mBAAmB,GAAG,EAAE,IAAI,CAAC,EAAE;QAC/D,gBAAgB,GAAG,GAAG,mBAAmB,IAAI,oBAAoB,GAAG,CAAA;KACpE;SAAM;QACN,gBAAgB,GAAG,YAAY,CAAA;KAC/B;IAED,OAAO,gBAAgB,CAAA;AACxB,CAAC"}
|
|
@@ -0,0 +1,6 @@
|
|
|
1
|
+
import { Timeline } from "../utilities/Timeline.js";
|
|
2
|
+
import { RawAudio } from "../audio/AudioUtilities.js";
|
|
3
|
+
export declare function recgonize(rawAudio: RawAudio, languageCode: string, region: string, accessKeyId: string, secretAccessKey: string): Promise<{
|
|
4
|
+
transcript: string;
|
|
5
|
+
timeline: Timeline;
|
|
6
|
+
}>;
|
|
@@ -0,0 +1,79 @@
|
|
|
1
|
+
import { wordCharacterPattern } from "../nlp/Segmentation.js";
|
|
2
|
+
import * as FFMpegTranscoder from "../codecs/FFMpegTranscoder.js";
|
|
3
|
+
import { Logger } from "../utilities/Logger.js";
|
|
4
|
+
export async function recgonize(rawAudio, languageCode, region, accessKeyId, secretAccessKey) {
|
|
5
|
+
const flac16Khz16bitMonoAudio = await FFMpegTranscoder.encodeFromChannels(rawAudio, { format: "flac", sampleRate: 16000, sampleFormat: "s16", channelCount: 1 });
|
|
6
|
+
const logger = new Logger();
|
|
7
|
+
logger.start("Initialize Amazon Transcribe streaming client module");
|
|
8
|
+
const streamingTranscribeSdk = await import("@aws-sdk/client-transcribe-streaming");
|
|
9
|
+
const streamingTranscribeClient = new streamingTranscribeSdk.TranscribeStreamingClient({
|
|
10
|
+
region,
|
|
11
|
+
credentials: {
|
|
12
|
+
accessKeyId,
|
|
13
|
+
secretAccessKey
|
|
14
|
+
}
|
|
15
|
+
});
|
|
16
|
+
const audioStream = async function* () {
|
|
17
|
+
const chunkSize = 2 ** 12;
|
|
18
|
+
//const audioSamples = encodeToAudioBuffer(rawAudio.audioChannels, 16, SampleFormat.PCM)
|
|
19
|
+
for (let i = 0; i < flac16Khz16bitMonoAudio.length; i += chunkSize) {
|
|
20
|
+
const chunk = flac16Khz16bitMonoAudio.subarray(i, i + chunkSize);
|
|
21
|
+
yield { AudioEvent: { AudioChunk: chunk } };
|
|
22
|
+
}
|
|
23
|
+
};
|
|
24
|
+
const params = {
|
|
25
|
+
LanguageCode: languageCode,
|
|
26
|
+
MediaSampleRateHertz: rawAudio.sampleRate,
|
|
27
|
+
MediaEncoding: "flac",
|
|
28
|
+
AudioStream: audioStream(),
|
|
29
|
+
};
|
|
30
|
+
logger.start("Request recognition from Amazon Transcribe..");
|
|
31
|
+
const command = new streamingTranscribeSdk.StartStreamTranscriptionCommand(params);
|
|
32
|
+
const response = await streamingTranscribeClient.send(command);
|
|
33
|
+
let transcript = "";
|
|
34
|
+
let events = [];
|
|
35
|
+
for await (const event of response.TranscriptResultStream) {
|
|
36
|
+
if (!event.TranscriptEvent) {
|
|
37
|
+
continue;
|
|
38
|
+
}
|
|
39
|
+
const transcriptEvent = event.TranscriptEvent;
|
|
40
|
+
const results = transcriptEvent.Transcript.Results;
|
|
41
|
+
if (results.length == 0) {
|
|
42
|
+
continue;
|
|
43
|
+
}
|
|
44
|
+
const firstResult = results[0];
|
|
45
|
+
const alternatives = firstResult.Alternatives;
|
|
46
|
+
if (!alternatives || alternatives.length == 0) {
|
|
47
|
+
continue;
|
|
48
|
+
}
|
|
49
|
+
const firstAlternative = alternatives[0];
|
|
50
|
+
//logger.log(firstAlternative.Transcript!)
|
|
51
|
+
if (firstResult.IsPartial === false) {
|
|
52
|
+
events = [...events, ...firstAlternative.Items];
|
|
53
|
+
transcript += " " + firstAlternative.Transcript;
|
|
54
|
+
}
|
|
55
|
+
}
|
|
56
|
+
logger.start("Process result");
|
|
57
|
+
transcript = transcript.replace(/ +/g, " ").trim();
|
|
58
|
+
const timeline = [];
|
|
59
|
+
for (const event of events) {
|
|
60
|
+
const text = event.Content;
|
|
61
|
+
if (!wordCharacterPattern.test(text)) {
|
|
62
|
+
continue;
|
|
63
|
+
}
|
|
64
|
+
const startTime = event.StartTime;
|
|
65
|
+
const endTime = event.EndTime;
|
|
66
|
+
const confidence = event.Confidence;
|
|
67
|
+
timeline[timeline.length - 1].endTime = startTime;
|
|
68
|
+
timeline.push({
|
|
69
|
+
type: "word",
|
|
70
|
+
text,
|
|
71
|
+
startTime,
|
|
72
|
+
endTime,
|
|
73
|
+
confidence
|
|
74
|
+
});
|
|
75
|
+
}
|
|
76
|
+
logger.end();
|
|
77
|
+
return { transcript, timeline };
|
|
78
|
+
}
|
|
79
|
+
//# sourceMappingURL=AmazonTranscribeSTT.js.map
|
|
@@ -0,0 +1 @@
|
|
|
1
|
+
{"version":3,"file":"AmazonTranscribeSTT.js","sourceRoot":"","sources":["../../src/recognition/AmazonTranscribeSTT.ts"],"names":[],"mappings":"AACA,OAAO,EAAE,oBAAoB,EAAE,MAAM,wBAAwB,CAAA;AAC7D,OAAO,KAAK,gBAAgB,MAAM,+BAA+B,CAAA;AACjE,OAAO,EAAE,MAAM,EAAE,MAAM,wBAAwB,CAAA;AAI/C,MAAM,CAAC,KAAK,UAAU,SAAS,CAAC,QAAkB,EAAE,YAAoB,EAAE,MAAc,EAAE,WAAmB,EAAE,eAAuB;IACrI,MAAM,uBAAuB,GAAG,MAAM,gBAAgB,CAAC,kBAAkB,CAAC,QAAQ,EAAE,EAAE,MAAM,EAAE,MAAM,EAAE,UAAU,EAAE,KAAK,EAAE,YAAY,EAAE,KAAK,EAAE,YAAY,EAAE,CAAC,EAAE,CAAC,CAAA;IAEhK,MAAM,MAAM,GAAG,IAAI,MAAM,EAAE,CAAA;IAC3B,MAAM,CAAC,KAAK,CAAC,sDAAsD,CAAC,CAAA;IAEpE,MAAM,sBAAsB,GAAG,MAAM,MAAM,CAAC,sCAAsC,CAAC,CAAA;IAEnF,MAAM,yBAAyB,GAAG,IAAI,sBAAsB,CAAC,yBAAyB,CAAC;QACtF,MAAM;QACN,WAAW,EAAE;YACZ,WAAW;YACX,eAAe;SACf;KACD,CAAC,CAAA;IAEF,MAAM,WAAW,GAAG,KAAK,SAAS,CAAC;QAClC,MAAM,SAAS,GAAG,CAAC,IAAI,EAAE,CAAA;QACzB,wFAAwF;QAExF,KAAK,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,GAAG,uBAAuB,CAAC,MAAM,EAAE,CAAC,IAAI,SAAS,EAAE;YACnE,MAAM,KAAK,GAAG,uBAAuB,CAAC,QAAQ,CAAC,CAAC,EAAE,CAAC,GAAG,SAAS,CAAC,CAAA;YAEhE,MAAM,EAAE,UAAU,EAAE,EAAE,UAAU,EAAE,KAAK,EAAE,EAAE,CAAA;SAC3C;IACF,CAAC,CAAA;IAED,MAAM,MAAM,GAAyC;QACpD,YAAY,EAAE,YAAY;QAC1B,oBAAoB,EAAE,QAAQ,CAAC,UAAU;QACzC,aAAa,EAAE,MAAM;QACrB,WAAW,EAAE,WAAW,EAAE;KAC1B,CAAA;IAED,MAAM,CAAC,KAAK,CAAC,8CAA8C,CAAC,CAAA;IAE5D,MAAM,OAAO,GAAG,IAAI,sBAAsB,CAAC,+BAA+B,CAAC,MAAM,CAAC,CAAA;IAElF,MAAM,QAAQ,GAAG,MAAM,yBAAyB,CAAC,IAAI,CAAC,OAAO,CAAC,CAAA;IAE9D,IAAI,UAAU,GAAG,EAAE,CAAA;IACnB,IAAI,MAAM,GAAW,EAAE,CAAA;IAEvB,IAAI,KAAK,EAAE,MAAM,KAAK,IAAI,QAAQ,CAAC,sBAAuB,EAAE;QAC3D,IAAI,CAAC,KAAK,CAAC,eAAe,EAAE;YAC3B,SAAQ;SACR;QAED,MAAM,eAAe,GAAG,KAAK,CAAC,eAAe,CAAA;QAE7C,MAAM,OAAO,GAAG,eAAe,CAAC,UAAW,CAAC,OAAQ,CAAA;QAEpD,IAAI,OAAO,CAAC,MAAM,IAAI,CAAC,EAAE;YACxB,SAAQ;SACR;QAED,MAAM,WAAW,GAAG,OAAO,CAAC,CAAC,CAAC,CAAA;QAC9B,MAAM,YAAY,GAAG,WAAW,CAAC,YAAY,CAAA;QAE7C,IAAI,CAAC,YAAY,IAAI,YAAY,CAAC,MAAM,IAAI,CAAC,EAAE;YAC9C,SAAQ;SACR;QAED,MAAM,gBAAgB,GAAG,YAAY,CAAC,CAAC,CAAC,CAAA;QACxC,0CAA0C;QAE1C,IAAI,WAAW,CAAC,SAAS,KAAK,KAAK,EAAE;YACpC,MAAM,GAAG,CAAC,GAAG,MAAM,EAAE,GAAG,gBAAgB,CAAC,KAAM,CAAC,CAAA;YAChD,UAAU,IAAI,GAAG,GAAG,gBAAgB,CAAC,UAAW,CAAA;SAChD;KACD;IAED,MAAM,CAAC,KAAK,CAAC,gBAAgB,CAAC,CAAA;IAE9B,UAAU,GAAG,UAAU,CAAC,OAAO,CAAC,KAAK,EAAE,GAAG,CAAC,CAAC,IAAI,EAAE,CAAA;IAElD,MAAM,QAAQ,GAAa,EAAE,CAAA;IAE7B,KAAK,MAAM,KAAK,IAAI,MAAM,EAAE;QAC3B,MAAM,IAAI,GAAG,KAAK,CAAC,OAAQ,CAAA;QAE3B,IAAI,CAAC,oBAAoB,CAAC,IAAI,CAAC,IAAI,CAAC,EAAE;YACrC,SAAQ;SACR;QAED,MAAM,SAAS,GAAG,KAAK,CAAC,SAAU,CAAA;QAClC,MAAM,OAAO,GAAG,KAAK,CAAC,OAAQ,CAAA;QAC9B,MAAM,UAAU,GAAG,KAAK,CAAC,UAAU,CAAA;QAEnC,QAAQ,CAAC,QAAQ,CAAC,MAAM,GAAG,CAAC,CAAC,CAAC,OAAO,GAAG,SAAS,CAAA;QAEjD,QAAQ,CAAC,IAAI,CACZ;YACC,IAAI,EAAE,MAAM;YACZ,IAAI;YACJ,SAAS;YACT,OAAO;YACP,UAAU;SACV,CAAC,CAAA;KACH;IAED,MAAM,CAAC,GAAG,EAAE,CAAA;IAEZ,OAAO,EAAE,UAAU,EAAE,QAAQ,EAAE,CAAA;AAChC,CAAC"}
|
|
@@ -0,0 +1,7 @@
|
|
|
1
|
+
import SpeechSDK from 'microsoft-cognitiveservices-speech-sdk';
|
|
2
|
+
import { RawAudio } from '../audio/AudioUtilities.js';
|
|
3
|
+
import { Timeline } from '../utilities/Timeline.js';
|
|
4
|
+
export declare function recognize(rawAudio: RawAudio, subscriptionKey: string, serviceRegion: string, languageCode: string, profanity?: SpeechSDK.ProfanityOption): Promise<{
|
|
5
|
+
transcript: string;
|
|
6
|
+
timeline: Timeline;
|
|
7
|
+
}>;
|
|
@@ -0,0 +1,51 @@
|
|
|
1
|
+
import SpeechSDK from 'microsoft-cognitiveservices-speech-sdk';
|
|
2
|
+
import { encodeWaveBuffer } from '../audio/AudioUtilities.js';
|
|
3
|
+
import { Logger } from '../utilities/Logger.js';
|
|
4
|
+
export async function recognize(rawAudio, subscriptionKey, serviceRegion, languageCode, profanity = SpeechSDK.ProfanityOption.Raw) {
|
|
5
|
+
const logger = new Logger();
|
|
6
|
+
logger.start("Request recognition from Azure Cognitive Services");
|
|
7
|
+
const result = await requestRecognition(rawAudio, subscriptionKey, serviceRegion, languageCode);
|
|
8
|
+
logger.start("Process result");
|
|
9
|
+
const transcript = result.text;
|
|
10
|
+
const resultObject = JSON.parse(result.json);
|
|
11
|
+
const bestResult = resultObject.NBest[0];
|
|
12
|
+
const timeline = [];
|
|
13
|
+
for (const wordEntry of bestResult.Words) {
|
|
14
|
+
const text = wordEntry.Word;
|
|
15
|
+
const startTime = wordEntry.Offset / 10000000;
|
|
16
|
+
const endTime = (wordEntry.Offset + wordEntry.Duration) / 10000000;
|
|
17
|
+
timeline.push({
|
|
18
|
+
type: "word",
|
|
19
|
+
text,
|
|
20
|
+
startTime,
|
|
21
|
+
endTime
|
|
22
|
+
});
|
|
23
|
+
}
|
|
24
|
+
logger.end();
|
|
25
|
+
return { transcript, timeline };
|
|
26
|
+
}
|
|
27
|
+
async function requestRecognition(rawAudio, subscriptionKey, serviceRegion, languageCode, profanity = SpeechSDK.ProfanityOption.Raw) {
|
|
28
|
+
//const encodedAudio = await FFMpegTranscoder.encodeFromChannels(rawAudio, { format: "wav", sampleRate: 16000, sampleFormat: "s16", channelCount: 1 });
|
|
29
|
+
const encodedAudio = encodeWaveBuffer(rawAudio);
|
|
30
|
+
return new Promise((resolve, reject) => {
|
|
31
|
+
const audioFormat = SpeechSDK.AudioStreamFormat.getWaveFormat(16000, 16, 1, SpeechSDK.AudioFormatTag.PCM);
|
|
32
|
+
const inputStream = SpeechSDK.AudioInputStream.createPushStream(audioFormat);
|
|
33
|
+
inputStream.write(encodedAudio);
|
|
34
|
+
inputStream.close();
|
|
35
|
+
const audioConfig = SpeechSDK.AudioConfig.fromStreamInput(inputStream);
|
|
36
|
+
const speechConfig = SpeechSDK.SpeechConfig.fromSubscription(subscriptionKey, serviceRegion);
|
|
37
|
+
speechConfig.speechRecognitionLanguage = languageCode;
|
|
38
|
+
speechConfig.setProfanity(profanity);
|
|
39
|
+
speechConfig.requestWordLevelTimestamps();
|
|
40
|
+
speechConfig.outputFormat = SpeechSDK.OutputFormat.Detailed;
|
|
41
|
+
const recognizer = new SpeechSDK.SpeechRecognizer(speechConfig, audioConfig);
|
|
42
|
+
recognizer.recognizeOnceAsync((result) => {
|
|
43
|
+
recognizer.close();
|
|
44
|
+
resolve(result);
|
|
45
|
+
}, (error) => {
|
|
46
|
+
recognizer.close();
|
|
47
|
+
reject(error);
|
|
48
|
+
});
|
|
49
|
+
});
|
|
50
|
+
}
|
|
51
|
+
//# sourceMappingURL=AzureCognitiveServicesSTT.js.map
|
|
@@ -0,0 +1 @@
|
|
|
1
|
+
{"version":3,"file":"AzureCognitiveServicesSTT.js","sourceRoot":"","sources":["../../src/recognition/AzureCognitiveServicesSTT.ts"],"names":[],"mappings":"AAAA,OAAO,SAAS,MAAM,wCAAwC,CAAA;AAE9D,OAAO,EAAY,gBAAgB,EAAE,MAAM,4BAA4B,CAAA;AACvE,OAAO,EAAE,MAAM,EAAE,MAAM,wBAAwB,CAAA;AAG/C,MAAM,CAAC,KAAK,UAAU,SAAS,CAAC,QAAkB,EAAE,eAAuB,EAAE,aAAqB,EAAE,YAAoB,EAAE,YAAuC,SAAS,CAAC,eAAe,CAAC,GAAG;IAC7L,MAAM,MAAM,GAAG,IAAI,MAAM,EAAE,CAAA;IAC3B,MAAM,CAAC,KAAK,CAAC,mDAAmD,CAAC,CAAA;IAEjE,MAAM,MAAM,GAAG,MAAM,kBAAkB,CAAC,QAAQ,EAAE,eAAe,EAAE,aAAa,EAAE,YAAY,CAAC,CAAA;IAE/F,MAAM,CAAC,KAAK,CAAC,gBAAgB,CAAC,CAAA;IAE9B,MAAM,UAAU,GAAG,MAAM,CAAC,IAAI,CAAA;IAE9B,MAAM,YAAY,GAAG,IAAI,CAAC,KAAK,CAAC,MAAM,CAAC,IAAI,CAAC,CAAA;IAC5C,MAAM,UAAU,GAAG,YAAY,CAAC,KAAK,CAAC,CAAC,CAAC,CAAA;IAExC,MAAM,QAAQ,GAAa,EAAE,CAAA;IAE7B,KAAK,MAAM,SAAS,IAAI,UAAU,CAAC,KAAK,EAAE;QACzC,MAAM,IAAI,GAAG,SAAS,CAAC,IAAI,CAAA;QAC3B,MAAM,SAAS,GAAG,SAAS,CAAC,MAAM,GAAG,QAAQ,CAAA;QAC7C,MAAM,OAAO,GAAG,CAAC,SAAS,CAAC,MAAM,GAAG,SAAS,CAAC,QAAQ,CAAC,GAAG,QAAQ,CAAA;QAElE,QAAQ,CAAC,IAAI,CAAC;YACb,IAAI,EAAE,MAAM;YACZ,IAAI;YACJ,SAAS;YACT,OAAO;SACP,CAAC,CAAA;KACF;IAED,MAAM,CAAC,GAAG,EAAE,CAAA;IAEZ,OAAO,EAAE,UAAU,EAAE,QAAQ,EAAE,CAAA;AAChC,CAAC;AAED,KAAK,UAAU,kBAAkB,CAAC,QAAkB,EAAE,eAAuB,EAAE,aAAqB,EAAE,YAAoB,EAAE,YAAuC,SAAS,CAAC,eAAe,CAAC,GAAG;IAC/L,uJAAuJ;IACvJ,MAAM,YAAY,GAAG,gBAAgB,CAAC,QAAQ,CAAC,CAAA;IAE/C,OAAO,IAAI,OAAO,CAAoC,CAAC,OAAO,EAAE,MAAM,EAAE,EAAE;QACzE,MAAM,WAAW,GAAG,SAAS,CAAC,iBAAiB,CAAC,aAAa,CAAC,KAAK,EAAE,EAAE,EAAE,CAAC,EAAE,SAAS,CAAC,cAAc,CAAC,GAAG,CAAC,CAAA;QAEzG,MAAM,WAAW,GAAG,SAAS,CAAC,gBAAgB,CAAC,gBAAgB,CAAC,WAAW,CAAC,CAAA;QAE5E,WAAW,CAAC,KAAK,CAAC,YAAY,CAAC,CAAA;QAC/B,WAAW,CAAC,KAAK,EAAE,CAAA;QAEnB,MAAM,WAAW,GAAG,SAAS,CAAC,WAAW,CAAC,eAAe,CAAC,WAAW,CAAC,CAAA;QAEtE,MAAM,YAAY,GAAG,SAAS,CAAC,YAAY,CAAC,gBAAgB,CAAC,eAAe,EAAE,aAAa,CAAC,CAAA;QAE5F,YAAY,CAAC,yBAAyB,GAAG,YAAY,CAAA;QAErD,YAAY,CAAC,YAAY,CAAC,SAAS,CAAC,CAAA;QACpC,YAAY,CAAC,0BAA0B,EAAE,CAAA;QAEzC,YAAY,CAAC,YAAY,GAAG,SAAS,CAAC,YAAY,CAAC,QAAQ,CAAA;QAE3D,MAAM,UAAU,GAAG,IAAI,SAAS,CAAC,gBAAgB,CAAC,YAAY,EAAE,WAAW,CAAC,CAAA;QAE5E,UAAU,CAAC,kBAAkB,CAC5B,CAAC,MAAM,EAAE,EAAE;YACV,UAAU,CAAC,KAAK,EAAE,CAAA;YAClB,OAAO,CAAC,MAAM,CAAC,CAAA;QAChB,CAAC,EAED,CAAC,KAAK,EAAE,EAAE;YACT,UAAU,CAAC,KAAK,EAAE,CAAA;YAClB,MAAM,CAAC,KAAK,CAAC,CAAA;QACd,CAAC,CAAC,CAAA;IACJ,CAAC,CAAC,CAAA;AACH,CAAC"}
|
|
@@ -0,0 +1,7 @@
|
|
|
1
|
+
import { Timeline } from "../utilities/Timeline.js";
|
|
2
|
+
import { RawAudio } from "../audio/AudioUtilities.js";
|
|
3
|
+
export type AudioEncoding = "LINEAR16" | "FLAC" | "MULAW" | "AMR" | "AMR" | "AMR_WB" | "OGG_OPUS" | "SPEEX_WITH_HEADER_BYTE" | "MP3" | "WEBM_OPUS";
|
|
4
|
+
export declare function recognize(rawAudio: RawAudio, apiKey: string, languageCode?: string): Promise<{
|
|
5
|
+
transcript: string;
|
|
6
|
+
timeline: Timeline;
|
|
7
|
+
}>;
|
|
@@ -0,0 +1,66 @@
|
|
|
1
|
+
import { request } from "gaxios";
|
|
2
|
+
import * as FFMpegTranscoder from "../codecs/FFMpegTranscoder.js";
|
|
3
|
+
import { Logger } from "../utilities/Logger.js";
|
|
4
|
+
export async function recognize(rawAudio, apiKey, languageCode = "en-US") {
|
|
5
|
+
const flac16Khz16bitMonoAudio = await FFMpegTranscoder.encodeFromChannels(rawAudio, { format: "flac", sampleRate: 16000, sampleFormat: "s16", channelCount: 1 });
|
|
6
|
+
const logger = new Logger();
|
|
7
|
+
logger.start("Request recognition from Google Cloud");
|
|
8
|
+
const requestBody = {
|
|
9
|
+
config: {
|
|
10
|
+
encoding: "FLAC",
|
|
11
|
+
sampleRateHertz: 16000,
|
|
12
|
+
audioChannelCount: 1,
|
|
13
|
+
languageCode,
|
|
14
|
+
alternativeLanguageCodes: [],
|
|
15
|
+
maxAlternatives: 1,
|
|
16
|
+
profanityFilter: false,
|
|
17
|
+
enableWordTimeOffsets: true,
|
|
18
|
+
enableWordConfidence: true,
|
|
19
|
+
enableAutomaticPunctuation: true,
|
|
20
|
+
model: "latest_long",
|
|
21
|
+
useEnhanced: true
|
|
22
|
+
},
|
|
23
|
+
audio: {
|
|
24
|
+
content: flac16Khz16bitMonoAudio.toString("base64")
|
|
25
|
+
}
|
|
26
|
+
};
|
|
27
|
+
const response = await request({
|
|
28
|
+
method: "POST",
|
|
29
|
+
url: `https://speech.googleapis.com/v1p1beta1/speech:recognize`,
|
|
30
|
+
params: {
|
|
31
|
+
"key": apiKey
|
|
32
|
+
},
|
|
33
|
+
headers: {
|
|
34
|
+
"User-Agent": ""
|
|
35
|
+
},
|
|
36
|
+
data: requestBody,
|
|
37
|
+
responseType: "json"
|
|
38
|
+
});
|
|
39
|
+
logger.start("Parse response body");
|
|
40
|
+
const result = parseResponseBody(response.data);
|
|
41
|
+
logger.end();
|
|
42
|
+
return result;
|
|
43
|
+
}
|
|
44
|
+
function parseResponseBody(responseBody) {
|
|
45
|
+
const results = responseBody.results;
|
|
46
|
+
let transcript = "";
|
|
47
|
+
const timeline = [];
|
|
48
|
+
for (const result of results) {
|
|
49
|
+
if (!result.alternatives || !result.alternatives[0] || !result.alternatives[0].transcript) {
|
|
50
|
+
continue;
|
|
51
|
+
}
|
|
52
|
+
const firstAlternative = result.alternatives[0];
|
|
53
|
+
transcript += firstAlternative.transcript;
|
|
54
|
+
for (const wordEvent of firstAlternative.words) {
|
|
55
|
+
timeline.push({
|
|
56
|
+
type: "word",
|
|
57
|
+
text: wordEvent.word,
|
|
58
|
+
startTime: parseFloat(wordEvent.startTime.replace("s", "")),
|
|
59
|
+
endTime: parseFloat(wordEvent.endTime.replace("s", "")),
|
|
60
|
+
confidence: wordEvent.confidence
|
|
61
|
+
});
|
|
62
|
+
}
|
|
63
|
+
}
|
|
64
|
+
return { transcript, timeline };
|
|
65
|
+
}
|
|
66
|
+
//# sourceMappingURL=GoogleCloudSTT.js.map
|
|
@@ -0,0 +1 @@
|
|
|
1
|
+
{"version":3,"file":"GoogleCloudSTT.js","sourceRoot":"","sources":["../../src/recognition/GoogleCloudSTT.ts"],"names":[],"mappings":"AAAA,OAAO,EAAE,OAAO,EAAE,MAAM,QAAQ,CAAA;AAEhC,OAAO,KAAK,gBAAgB,MAAM,+BAA+B,CAAA;AACjE,OAAO,EAAE,MAAM,EAAE,MAAM,wBAAwB,CAAA;AAM/C,MAAM,CAAC,KAAK,UAAU,SAAS,CAAC,QAAkB,EAAE,MAAc,EAAE,YAAY,GAAG,OAAO;IACzF,MAAM,uBAAuB,GAAG,MAAM,gBAAgB,CAAC,kBAAkB,CAAC,QAAQ,EAAE,EAAE,MAAM,EAAE,MAAM,EAAE,UAAU,EAAE,KAAK,EAAE,YAAY,EAAE,KAAK,EAAE,YAAY,EAAE,CAAC,EAAE,CAAC,CAAA;IAEhK,MAAM,MAAM,GAAG,IAAI,MAAM,EAAE,CAAA;IAC3B,MAAM,CAAC,KAAK,CAAC,uCAAuC,CAAC,CAAA;IAErD,MAAM,WAAW,GAAG;QACnB,MAAM,EAAE;YACP,QAAQ,EAAE,MAAM;YAChB,eAAe,EAAE,KAAK;YACtB,iBAAiB,EAAE,CAAC;YACpB,YAAY;YACZ,wBAAwB,EAAE,EAAE;YAC5B,eAAe,EAAE,CAAC;YAClB,eAAe,EAAE,KAAK;YACtB,qBAAqB,EAAE,IAAI;YAC3B,oBAAoB,EAAE,IAAI;YAC1B,0BAA0B,EAAE,IAAI;YAChC,KAAK,EAAE,aAAa;YACpB,WAAW,EAAE,IAAI;SACjB;QAED,KAAK,EAAE;YACN,OAAO,EAAE,uBAAuB,CAAC,QAAQ,CAAC,QAAQ,CAAC;SACnD;KACD,CAAA;IAED,MAAM,QAAQ,GAAG,MAAM,OAAO,CAAM;QACnC,MAAM,EAAE,MAAM;QAEd,GAAG,EAAE,0DAA0D;QAE/D,MAAM,EAAE;YACP,KAAK,EAAE,MAAM;SACb;QAED,OAAO,EAAE;YACR,YAAY,EAAE,EAAE;SAChB;QAED,IAAI,EAAE,WAAW;QAGjB,YAAY,EAAE,MAAM;KACpB,CAAC,CAAA;IAEF,MAAM,CAAC,KAAK,CAAC,qBAAqB,CAAC,CAAA;IAEnC,MAAM,MAAM,GAAG,iBAAiB,CAAC,QAAQ,CAAC,IAAI,CAAC,CAAA;IAE/C,MAAM,CAAC,GAAG,EAAE,CAAA;IAEZ,OAAO,MAAM,CAAA;AACd,CAAC;AAED,SAAS,iBAAiB,CAAC,YAAiB;IAC3C,MAAM,OAAO,GAAG,YAAY,CAAC,OAAO,CAAA;IAEpC,IAAI,UAAU,GAAG,EAAE,CAAA;IACnB,MAAM,QAAQ,GAAa,EAAE,CAAA;IAE7B,KAAK,MAAM,MAAM,IAAI,OAAO,EAAE;QAC7B,IAAI,CAAC,MAAM,CAAC,YAAY,IAAI,CAAC,MAAM,CAAC,YAAY,CAAC,CAAC,CAAC,IAAI,CAAC,MAAM,CAAC,YAAY,CAAC,CAAC,CAAC,CAAC,UAAU,EAAE;YAC1F,SAAQ;SACR;QAED,MAAM,gBAAgB,GAAG,MAAM,CAAC,YAAY,CAAC,CAAC,CAAC,CAAA;QAE/C,UAAU,IAAI,gBAAgB,CAAC,UAAU,CAAA;QAEzC,KAAK,MAAM,SAAS,IAAI,gBAAgB,CAAC,KAAK,EAAE;YAC/C,QAAQ,CAAC,IAAI,CAAC;gBACb,IAAI,EAAE,MAAM;gBACZ,IAAI,EAAE,SAAS,CAAC,IAAI;gBACpB,SAAS,EAAE,UAAU,CAAC,SAAS,CAAC,SAAS,CAAC,OAAO,CAAC,GAAG,EAAC,EAAE,CAAC,CAAC;gBAC1D,OAAO,EAAE,UAAU,CAAC,SAAS,CAAC,OAAO,CAAC,OAAO,CAAC,GAAG,EAAE,EAAE,CAAC,CAAC;gBACvD,UAAU,EAAE,SAAS,CAAC,UAAU;aAChC,CAAC,CAAA;SACF;KACD;IAED,OAAO,EAAE,UAAU,EAAE,QAAQ,EAAE,CAAA;AAChC,CAAC"}
|
|
@@ -0,0 +1,9 @@
|
|
|
1
|
+
import { Timeline } from "../utilities/Timeline.js";
|
|
2
|
+
import { RawAudio } from "../audio/AudioUtilities.js";
|
|
3
|
+
export declare function recognize(rawAudio: RawAudio, modelDirectory: string): Promise<{
|
|
4
|
+
transcript: string;
|
|
5
|
+
timeline: Timeline;
|
|
6
|
+
}>;
|
|
7
|
+
export declare const languageCodeToPackageName: {
|
|
8
|
+
[languageCode: string]: string;
|
|
9
|
+
};
|
|
@@ -0,0 +1,125 @@
|
|
|
1
|
+
import { indexOfMax } from "../math/VectorMath.js";
|
|
2
|
+
import { wordCharacterPattern } from '../nlp/Segmentation.js';
|
|
3
|
+
import Onnx from 'onnxruntime-node';
|
|
4
|
+
import { Logger } from '../utilities/Logger.js';
|
|
5
|
+
import { logToStderr } from "../utilities/Utilities.js";
|
|
6
|
+
import { getRawAudioDuration } from "../audio/AudioUtilities.js";
|
|
7
|
+
import { readAndParseJsonFile } from "../utilities/FileSystem.js";
|
|
8
|
+
import path from "path";
|
|
9
|
+
const log = logToStderr;
|
|
10
|
+
export async function recognize(rawAudio, modelDirectory) {
|
|
11
|
+
const logger = new Logger();
|
|
12
|
+
logger.start("Create ONNX inference session");
|
|
13
|
+
const modelPath = path.join(modelDirectory, "model.onnx");
|
|
14
|
+
const labelsPath = path.join(modelDirectory, "labels.json");
|
|
15
|
+
const labels = await readAndParseJsonFile(labelsPath);
|
|
16
|
+
const onnxOptions = {
|
|
17
|
+
logSeverityLevel: 3
|
|
18
|
+
};
|
|
19
|
+
const recognition = await Onnx.InferenceSession.create(modelPath, onnxOptions);
|
|
20
|
+
logger.start("Prepare input data");
|
|
21
|
+
const audioSamples = rawAudio.audioChannels[0];
|
|
22
|
+
const inputTensor = new Onnx.Tensor('float32', audioSamples, [1, audioSamples.length]);
|
|
23
|
+
const inputs = { input: inputTensor };
|
|
24
|
+
logger.start("Recognize with silero model");
|
|
25
|
+
const results = await recognition.run(inputs);
|
|
26
|
+
const rawResultValues = results["output"].data;
|
|
27
|
+
const tokenResults = [];
|
|
28
|
+
for (let i = 0; i < rawResultValues.length; i += labels.length) {
|
|
29
|
+
tokenResults.push(rawResultValues.subarray(i, i + labels.length));
|
|
30
|
+
}
|
|
31
|
+
const tokens = [];
|
|
32
|
+
for (const tokenResult of tokenResults) {
|
|
33
|
+
const bestCandidateIndex = indexOfMax(new Array(...tokenResult));
|
|
34
|
+
tokens.push(labels[bestCandidateIndex]);
|
|
35
|
+
}
|
|
36
|
+
//log(tokens.join("|"))
|
|
37
|
+
const result = processTokens(tokens, getRawAudioDuration(rawAudio));
|
|
38
|
+
logger.end();
|
|
39
|
+
return result;
|
|
40
|
+
}
|
|
41
|
+
function processTokens(tokens, totalDuration) {
|
|
42
|
+
const tokenCount = tokens.length;
|
|
43
|
+
const decodedTokens = [];
|
|
44
|
+
let tokenGroupIndexes = [[]];
|
|
45
|
+
for (let i = 0; i < tokenCount; i++) {
|
|
46
|
+
const token = tokens[i];
|
|
47
|
+
if (token == "2") {
|
|
48
|
+
if (decodedTokens.length > 0) {
|
|
49
|
+
const previousDecodedToken = decodedTokens[decodedTokens.length - 1];
|
|
50
|
+
decodedTokens.push("$");
|
|
51
|
+
decodedTokens.push(previousDecodedToken);
|
|
52
|
+
tokenGroupIndexes[tokenGroupIndexes.length - 1].push(i);
|
|
53
|
+
}
|
|
54
|
+
else {
|
|
55
|
+
decodedTokens.push(" ");
|
|
56
|
+
tokenGroupIndexes.push([]);
|
|
57
|
+
}
|
|
58
|
+
continue;
|
|
59
|
+
}
|
|
60
|
+
if (token == "_") {
|
|
61
|
+
continue;
|
|
62
|
+
}
|
|
63
|
+
decodedTokens.push(token);
|
|
64
|
+
if (token == " ") {
|
|
65
|
+
tokenGroupIndexes.push([]);
|
|
66
|
+
}
|
|
67
|
+
else {
|
|
68
|
+
tokenGroupIndexes[tokenGroupIndexes.length - 1].push(i);
|
|
69
|
+
}
|
|
70
|
+
}
|
|
71
|
+
let decodedString = "";
|
|
72
|
+
for (let i = 0; i < decodedTokens.length; i++) {
|
|
73
|
+
const currentToken = decodedTokens[i];
|
|
74
|
+
const previousToken = decodedTokens[i - 1];
|
|
75
|
+
if (currentToken != "$" && (previousToken != currentToken || previousToken == undefined)) {
|
|
76
|
+
decodedString += currentToken;
|
|
77
|
+
}
|
|
78
|
+
}
|
|
79
|
+
decodedString = decodedString.trim();
|
|
80
|
+
tokenGroupIndexes = tokenGroupIndexes.filter(group => group.length > 0);
|
|
81
|
+
if (tokenGroupIndexes.length > 0) {
|
|
82
|
+
let currentCorrection = Math.min(tokenGroupIndexes[0][0], 1.5);
|
|
83
|
+
for (let i = 0; i < tokenGroupIndexes.length; i++) {
|
|
84
|
+
const group = tokenGroupIndexes[i];
|
|
85
|
+
if (group.length == 1) {
|
|
86
|
+
group.push(group[0]);
|
|
87
|
+
}
|
|
88
|
+
group[0] -= currentCorrection;
|
|
89
|
+
if (i == tokenGroupIndexes.length - 1) {
|
|
90
|
+
currentCorrection = Math.min(tokenCount - i, 1.5);
|
|
91
|
+
}
|
|
92
|
+
else {
|
|
93
|
+
currentCorrection = Math.min((tokenGroupIndexes[i + 1][0] - group[group.length - 1]) / 2, 1.5);
|
|
94
|
+
}
|
|
95
|
+
group[group.length - 1] += currentCorrection;
|
|
96
|
+
}
|
|
97
|
+
}
|
|
98
|
+
const words = decodedString.split(" ");
|
|
99
|
+
const timeMultiplier = totalDuration / tokenCount;
|
|
100
|
+
const timeline = [];
|
|
101
|
+
for (let i = 0; i < words.length; i++) {
|
|
102
|
+
const text = words[i];
|
|
103
|
+
if (!wordCharacterPattern.test(text)) {
|
|
104
|
+
continue;
|
|
105
|
+
}
|
|
106
|
+
const group = tokenGroupIndexes[i];
|
|
107
|
+
const startTime = group[0] * timeMultiplier;
|
|
108
|
+
const endTime = group[group.length - 1] * timeMultiplier;
|
|
109
|
+
timeline.push({
|
|
110
|
+
type: "word",
|
|
111
|
+
text: text,
|
|
112
|
+
startTime,
|
|
113
|
+
endTime,
|
|
114
|
+
});
|
|
115
|
+
}
|
|
116
|
+
timeline[timeline.length - 1].endTime = totalDuration;
|
|
117
|
+
return { transcript: decodedString, timeline };
|
|
118
|
+
}
|
|
119
|
+
export const languageCodeToPackageName = {
|
|
120
|
+
"en": "silero-en-v5",
|
|
121
|
+
"es": "silero-es-v1",
|
|
122
|
+
"de": "silero-de-v1",
|
|
123
|
+
"uk": "silero-ua-v3",
|
|
124
|
+
};
|
|
125
|
+
//# sourceMappingURL=SileroSTT.js.map
|
|
@@ -0,0 +1 @@
|
|
|
1
|
+
{"version":3,"file":"SileroSTT.js","sourceRoot":"","sources":["../../src/recognition/SileroSTT.ts"],"names":[],"mappings":"AAAA,OAAO,EAAE,UAAU,EAAE,MAAM,uBAAuB,CAAA;AAClD,OAAO,EAAE,oBAAoB,EAAE,MAAM,wBAAwB,CAAA;AAC7D,OAAO,IAAI,MAAM,kBAAkB,CAAA;AACnC,OAAO,EAAE,MAAM,EAAE,MAAM,wBAAwB,CAAA;AAC/C,OAAO,EAAE,WAAW,EAAE,MAAM,2BAA2B,CAAA;AAEvD,OAAO,EAAY,mBAAmB,EAAE,MAAM,4BAA4B,CAAA;AAC1E,OAAO,EAAE,oBAAoB,EAAY,MAAM,4BAA4B,CAAA;AAC3E,OAAO,IAAI,MAAM,MAAM,CAAA;AAEvB,MAAM,GAAG,GAAG,WAAW,CAAA;AAEvB,MAAM,CAAC,KAAK,UAAU,SAAS,CAAC,QAAkB,EAAE,cAAsB;IACzE,MAAM,MAAM,GAAG,IAAI,MAAM,EAAE,CAAA;IAC3B,MAAM,CAAC,KAAK,CAAC,+BAA+B,CAAC,CAAA;IAE7C,MAAM,SAAS,GAAG,IAAI,CAAC,IAAI,CAAC,cAAc,EAAE,YAAY,CAAC,CAAA;IACzD,MAAM,UAAU,GAAG,IAAI,CAAC,IAAI,CAAC,cAAc,EAAE,aAAa,CAAC,CAAA;IAE3D,MAAM,MAAM,GAAa,MAAM,oBAAoB,CAAC,UAAU,CAAC,CAAA;IAE/D,MAAM,WAAW,GAAyC;QACzD,gBAAgB,EAAE,CAAC;KACnB,CAAA;IAED,MAAM,WAAW,GAAG,MAAM,IAAI,CAAC,gBAAgB,CAAC,MAAM,CAAC,SAAS,EAAE,WAAW,CAAC,CAAA;IAE9E,MAAM,CAAC,KAAK,CAAC,oBAAoB,CAAC,CAAA;IAElC,MAAM,YAAY,GAAG,QAAQ,CAAC,aAAa,CAAC,CAAC,CAAC,CAAA;IAE9C,MAAM,WAAW,GAAG,IAAI,IAAI,CAAC,MAAM,CAAC,SAAS,EAAE,YAAY,EAAE,CAAC,CAAC,EAAE,YAAY,CAAC,MAAM,CAAC,CAAC,CAAA;IAEtF,MAAM,MAAM,GAAG,EAAE,KAAK,EAAE,WAAW,EAAE,CAAA;IAErC,MAAM,CAAC,KAAK,CAAC,6BAA6B,CAAC,CAAA;IAE3C,MAAM,OAAO,GAAG,MAAM,WAAW,CAAC,GAAG,CAAC,MAAM,CAAC,CAAA;IAE7C,MAAM,eAAe,GAAG,OAAO,CAAC,QAAQ,CAAC,CAAC,IAAoB,CAAA;IAE9D,MAAM,YAAY,GAAmB,EAAE,CAAA;IAEvC,KAAK,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,GAAG,eAAe,CAAC,MAAM,EAAE,CAAC,IAAI,MAAM,CAAC,MAAM,EAAE;QAC/D,YAAY,CAAC,IAAI,CAAC,eAAe,CAAC,QAAQ,CAAC,CAAC,EAAE,CAAC,GAAG,MAAM,CAAC,MAAM,CAAC,CAAC,CAAA;KACjE;IAED,MAAM,MAAM,GAAa,EAAE,CAAA;IAE3B,KAAK,MAAM,WAAW,IAAI,YAAY,EAAE;QACvC,MAAM,kBAAkB,GAAG,UAAU,CAAC,IAAI,KAAK,CAAC,GAAG,WAAW,CAAC,CAAC,CAAA;QAChE,MAAM,CAAC,IAAI,CAAC,MAAM,CAAC,kBAAkB,CAAC,CAAC,CAAA;KACvC;IAED,uBAAuB;IAEvB,MAAM,MAAM,GAAG,aAAa,CAAC,MAAM,EAAE,mBAAmB,CAAC,QAAQ,CAAC,CAAC,CAAA;IAEnE,MAAM,CAAC,GAAG,EAAE,CAAA;IAEZ,OAAO,MAAM,CAAA;AACd,CAAC;AAED,SAAS,aAAa,CAAC,MAAgB,EAAE,aAAqB;IAC7D,MAAM,UAAU,GAAG,MAAM,CAAC,MAAM,CAAA;IAEhC,MAAM,aAAa,GAAa,EAAE,CAAA;IAClC,IAAI,iBAAiB,GAAe,CAAC,EAAE,CAAC,CAAA;IAExC,KAAK,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,GAAG,UAAU,EAAE,CAAC,EAAE,EAAE;QACpC,MAAM,KAAK,GAAG,MAAM,CAAC,CAAC,CAAC,CAAA;QAEvB,IAAI,KAAK,IAAI,GAAG,EAAE;YACjB,IAAI,aAAa,CAAC,MAAM,GAAG,CAAC,EAAE;gBAC7B,MAAM,oBAAoB,GAAG,aAAa,CAAC,aAAa,CAAC,MAAM,GAAG,CAAC,CAAC,CAAA;gBACpE,aAAa,CAAC,IAAI,CAAC,GAAG,CAAC,CAAA;gBACvB,aAAa,CAAC,IAAI,CAAC,oBAAoB,CAAC,CAAA;gBAExC,iBAAiB,CAAC,iBAAiB,CAAC,MAAM,GAAG,CAAC,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC,CAAA;aACvD;iBAAM;gBACN,aAAa,CAAC,IAAI,CAAC,GAAG,CAAC,CAAA;gBACvB,iBAAiB,CAAC,IAAI,CAAC,EAAE,CAAC,CAAA;aAC1B;YAED,SAAQ;SACR;QAED,IAAI,KAAK,IAAI,GAAG,EAAE;YACjB,SAAQ;SACR;QAED,aAAa,CAAC,IAAI,CAAC,KAAK,CAAC,CAAA;QAEzB,IAAI,KAAK,IAAI,GAAG,EAAE;YACjB,iBAAiB,CAAC,IAAI,CAAC,EAAE,CAAC,CAAA;SAC1B;aAAM;YACN,iBAAiB,CAAC,iBAAiB,CAAC,MAAM,GAAG,CAAC,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC,CAAA;SACvD;KACD;IAED,IAAI,aAAa,GAAG,EAAE,CAAA;IAEtB,KAAK,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,GAAG,aAAa,CAAC,MAAM,EAAE,CAAC,EAAE,EAAE;QAC9C,MAAM,YAAY,GAAG,aAAa,CAAC,CAAC,CAAC,CAAA;QACrC,MAAM,aAAa,GAAG,aAAa,CAAC,CAAC,GAAG,CAAC,CAAC,CAAA;QAE1C,IAAI,YAAY,IAAI,GAAG,IAAI,CAAC,aAAa,IAAI,YAAY,IAAI,aAAa,IAAI,SAAS,CAAC,EAAE;YACzF,aAAa,IAAI,YAAY,CAAA;SAC7B;KACD;IAED,aAAa,GAAG,aAAa,CAAC,IAAI,EAAE,CAAA;IAEpC,iBAAiB,GAAG,iBAAiB,CAAC,MAAM,CAAC,KAAK,CAAC,EAAE,CAAC,KAAK,CAAC,MAAM,GAAG,CAAC,CAAC,CAAA;IAEvE,IAAI,iBAAiB,CAAC,MAAM,GAAG,CAAC,EAAE;QACjC,IAAI,iBAAiB,GAAG,IAAI,CAAC,GAAG,CAAC,iBAAiB,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,EAAE,GAAG,CAAC,CAAA;QAE9D,KAAK,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,GAAG,iBAAiB,CAAC,MAAM,EAAE,CAAC,EAAE,EAAE;YAClD,MAAM,KAAK,GAAG,iBAAiB,CAAC,CAAC,CAAC,CAAA;YAElC,IAAI,KAAK,CAAC,MAAM,IAAI,CAAC,EAAE;gBACtB,KAAK,CAAC,IAAI,CAAC,KAAK,CAAC,CAAC,CAAC,CAAC,CAAA;aACpB;YAED,KAAK,CAAC,CAAC,CAAC,IAAI,iBAAiB,CAAA;YAE7B,IAAI,CAAC,IAAI,iBAAiB,CAAC,MAAM,GAAG,CAAC,EAAE;gBACtC,iBAAiB,GAAG,IAAI,CAAC,GAAG,CAAC,UAAU,GAAG,CAAC,EAAE,GAAG,CAAC,CAAA;aACjD;iBAAM;gBACN,iBAAiB,GAAG,IAAI,CAAC,GAAG,CAAC,CAAC,iBAAiB,CAAC,CAAC,GAAG,CAAC,CAAC,CAAC,CAAC,CAAC,GAAG,KAAK,CAAC,KAAK,CAAC,MAAM,GAAG,CAAC,CAAC,CAAC,GAAG,CAAC,EAAE,GAAG,CAAC,CAAA;aAC9F;YAED,KAAK,CAAC,KAAK,CAAC,MAAM,GAAG,CAAC,CAAC,IAAI,iBAAiB,CAAA;SAC5C;KACD;IAED,MAAM,KAAK,GAAG,aAAa,CAAC,KAAK,CAAC,GAAG,CAAC,CAAA;IAEtC,MAAM,cAAc,GAAG,aAAa,GAAG,UAAU,CAAA;IAEjD,MAAM,QAAQ,GAAa,EAAE,CAAA;IAE7B,KAAK,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,GAAG,KAAK,CAAC,MAAM,EAAE,CAAC,EAAE,EAAE;QACtC,MAAM,IAAI,GAAG,KAAK,CAAC,CAAC,CAAC,CAAA;QAErB,IAAI,CAAC,oBAAoB,CAAC,IAAI,CAAC,IAAI,CAAC,EAAE;YACrC,SAAQ;SACR;QAED,MAAM,KAAK,GAAG,iBAAiB,CAAC,CAAC,CAAC,CAAA;QAClC,MAAM,SAAS,GAAG,KAAK,CAAC,CAAC,CAAC,GAAG,cAAc,CAAA;QAC3C,MAAM,OAAO,GAAG,KAAK,CAAC,KAAK,CAAC,MAAM,GAAG,CAAC,CAAC,GAAG,cAAc,CAAA;QAExD,QAAQ,CAAC,IAAI,CAAC;YACb,IAAI,EAAE,MAAM;YACZ,IAAI,EAAE,IAAI;YACV,SAAS;YACT,OAAO;SACP,CAAC,CAAA;KACF;IAED,QAAQ,CAAC,QAAQ,CAAC,MAAM,GAAG,CAAC,CAAC,CAAC,OAAO,GAAG,aAAa,CAAA;IAErD,OAAO,EAAE,UAAU,EAAE,aAAa,EAAE,QAAQ,EAAE,CAAA;AAC/C,CAAC;AAED,MAAM,CAAC,MAAM,yBAAyB,GAAuC;IAC5E,IAAI,EAAE,cAAc;IACpB,IAAI,EAAE,cAAc;IACpB,IAAI,EAAE,cAAc;IACpB,IAAI,EAAE,cAAc;CACpB,CAAA"}
|
|
@@ -0,0 +1,10 @@
|
|
|
1
|
+
import { Timeline } from "../utilities/Timeline.js";
|
|
2
|
+
import { RawAudio } from "../audio/AudioUtilities.js";
|
|
3
|
+
export declare function recognizeFile(filename: string, modelPath: string, verbose?: boolean): Promise<{
|
|
4
|
+
transcript: string;
|
|
5
|
+
timeline: Timeline;
|
|
6
|
+
}>;
|
|
7
|
+
export declare function recognize(rawAudio: RawAudio, modelPath: string, verbose?: boolean): Promise<{
|
|
8
|
+
transcript: string;
|
|
9
|
+
timeline: Timeline;
|
|
10
|
+
}>;
|
|
@@ -0,0 +1,78 @@
|
|
|
1
|
+
import * as FFMpegTranscoder from "../codecs/FFMpegTranscoder.js";
|
|
2
|
+
import * as AudioBufferConversion from '../audio/AudioBufferConversion.js';
|
|
3
|
+
import { Logger } from "../utilities/Logger.js";
|
|
4
|
+
import { logToStderr } from "../utilities/Utilities.js";
|
|
5
|
+
const log = logToStderr;
|
|
6
|
+
export async function recognizeFile(filename, modelPath, verbose = true) {
|
|
7
|
+
const rawAudio = await FFMpegTranscoder.decodeToChannels(filename, 16000, 1);
|
|
8
|
+
return recognize(rawAudio, modelPath, verbose);
|
|
9
|
+
}
|
|
10
|
+
export async function recognize(rawAudio, modelPath, verbose = true) {
|
|
11
|
+
const logger = new Logger();
|
|
12
|
+
logger.start("Initialize vosk recognizer");
|
|
13
|
+
const audioChannels = rawAudio.audioChannels;
|
|
14
|
+
const sampleRate = rawAudio.sampleRate;
|
|
15
|
+
let Vosk;
|
|
16
|
+
try {
|
|
17
|
+
Vosk = await import('@echogarden/vosk');
|
|
18
|
+
}
|
|
19
|
+
catch (e) {
|
|
20
|
+
throw new Error(`vosk package, which is required for Vosk support, was not found. You can install it by running 'npm install @echogarden/vosk -g'.`);
|
|
21
|
+
}
|
|
22
|
+
Vosk.setLogLevel(-1);
|
|
23
|
+
const model = await new Vosk.Model(modelPath);
|
|
24
|
+
const recognizer = new Vosk.Recognizer({ model, sampleRate });
|
|
25
|
+
recognizer.setMaxAlternatives(0);
|
|
26
|
+
recognizer.setWords(true);
|
|
27
|
+
recognizer.setPartialWords(true);
|
|
28
|
+
logger.start("Recognize with vosk");
|
|
29
|
+
const recognitionStartTimestamp = logger.getTimestamp();
|
|
30
|
+
const pcmAudio = AudioBufferConversion.encodeToAudioBuffer(audioChannels, 16);
|
|
31
|
+
const trailingSilence = Buffer.alloc(sampleRate * 4);
|
|
32
|
+
const pcmAudioWithTrailingSilence = Buffer.concat([pcmAudio, trailingSilence]);
|
|
33
|
+
const pcmAudioByteCount = pcmAudioWithTrailingSilence.length;
|
|
34
|
+
const maxChunkSize = sampleRate * 2.0;
|
|
35
|
+
let previousResultText = "";
|
|
36
|
+
for (let readOffset = 0; readOffset < pcmAudioByteCount; readOffset += maxChunkSize) {
|
|
37
|
+
const chunkSize = Math.min(maxChunkSize, pcmAudioByteCount - readOffset);
|
|
38
|
+
const chunk = pcmAudioWithTrailingSilence.subarray(readOffset, readOffset + chunkSize);
|
|
39
|
+
const speechEnded = await recognizer.acceptWaveformAsync(chunk);
|
|
40
|
+
if (verbose) {
|
|
41
|
+
const partialResultText = recognizer.partialResult().partial;
|
|
42
|
+
if (partialResultText != previousResultText) {
|
|
43
|
+
//logger.log(partialResultText)
|
|
44
|
+
//logger.log("")
|
|
45
|
+
previousResultText = partialResultText;
|
|
46
|
+
}
|
|
47
|
+
}
|
|
48
|
+
}
|
|
49
|
+
//const speechEnded = await recognizer.acceptWaveformAsync(pcmAudioWithTrailingSilence)
|
|
50
|
+
const result = recognizer.finalResult();
|
|
51
|
+
recognizer.reset();
|
|
52
|
+
recognizer.free();
|
|
53
|
+
model.free();
|
|
54
|
+
const transcript = result.text;
|
|
55
|
+
const events = result.result;
|
|
56
|
+
if (events.length == 0) {
|
|
57
|
+
return { transcript, timeline: [] };
|
|
58
|
+
}
|
|
59
|
+
const timeline = [];
|
|
60
|
+
for (let i = 0; i < events.length; i++) {
|
|
61
|
+
const event = events[i];
|
|
62
|
+
const eventText = event.word;
|
|
63
|
+
const eventStart = event.start;
|
|
64
|
+
const eventEnd = event.end;
|
|
65
|
+
const eventConfidence = event.conf;
|
|
66
|
+
timeline.push({
|
|
67
|
+
type: "word",
|
|
68
|
+
text: eventText,
|
|
69
|
+
startTime: eventStart,
|
|
70
|
+
endTime: eventEnd,
|
|
71
|
+
confidence: eventConfidence
|
|
72
|
+
});
|
|
73
|
+
}
|
|
74
|
+
//logger.logDuration(`Recognition with vosk`, recognitionStartTimestamp)
|
|
75
|
+
logger.end();
|
|
76
|
+
return { transcript, timeline };
|
|
77
|
+
}
|
|
78
|
+
//# sourceMappingURL=VoskSTT.js.map
|
|
@@ -0,0 +1 @@
|
|
|
1
|
+
{"version":3,"file":"VoskSTT.js","sourceRoot":"","sources":["../../src/recognition/VoskSTT.ts"],"names":[],"mappings":"AAAA,OAAO,KAAK,gBAAgB,MAAM,+BAA+B,CAAA;AACjE,OAAO,KAAK,qBAAqB,MAAM,mCAAmC,CAAA;AAC1E,OAAO,EAAE,MAAM,EAAE,MAAM,wBAAwB,CAAA;AAC/C,OAAO,EAAE,WAAW,EAAE,MAAM,2BAA2B,CAAA;AAIvD,MAAM,GAAG,GAAG,WAAW,CAAA;AAEvB,MAAM,CAAC,KAAK,UAAU,aAAa,CAAC,QAAgB,EAAE,SAAiB,EAAE,OAAO,GAAG,IAAI;IACtF,MAAM,QAAQ,GAAG,MAAM,gBAAgB,CAAC,gBAAgB,CAAC,QAAQ,EAAE,KAAK,EAAE,CAAC,CAAC,CAAA;IAC5E,OAAO,SAAS,CAAC,QAAQ,EAAE,SAAS,EAAE,OAAO,CAAC,CAAA;AAC/C,CAAC;AAED,MAAM,CAAC,KAAK,UAAU,SAAS,CAAC,QAAkB,EAAE,SAAiB,EAAE,OAAO,GAAG,IAAI;IACpF,MAAM,MAAM,GAAG,IAAI,MAAM,EAAE,CAAA;IAC3B,MAAM,CAAC,KAAK,CAAC,4BAA4B,CAAC,CAAA;IAE1C,MAAM,aAAa,GAAG,QAAQ,CAAC,aAAa,CAAA;IAC5C,MAAM,UAAU,GAAG,QAAQ,CAAC,UAAU,CAAA;IAEtC,IAAI,IAAS,CAAA;IAEb,IAAI;QACH,IAAI,GAAG,MAAM,MAAM,CAAC,kBAAkB,CAAC,CAAA;KACvC;IAAC,OAAO,CAAC,EAAE;QACX,MAAM,IAAI,KAAK,CAAC,mIAAmI,CAAC,CAAA;KACpJ;IAED,IAAI,CAAC,WAAW,CAAC,CAAC,CAAC,CAAC,CAAA;IAEpB,MAAM,KAAK,GAAG,MAAM,IAAI,IAAI,CAAC,KAAK,CAAC,SAAS,CAAC,CAAA;IAE7C,MAAM,UAAU,GAAG,IAAI,IAAI,CAAC,UAAU,CAAC,EAAE,KAAK,EAAE,UAAU,EAAE,CAAC,CAAA;IAE7D,UAAU,CAAC,kBAAkB,CAAC,CAAC,CAAC,CAAA;IAChC,UAAU,CAAC,QAAQ,CAAC,IAAI,CAAC,CAAA;IACzB,UAAU,CAAC,eAAe,CAAC,IAAI,CAAC,CAAA;IAEhC,MAAM,CAAC,KAAK,CAAC,qBAAqB,CAAC,CAAA;IAEnC,MAAM,yBAAyB,GAAG,MAAM,CAAC,YAAY,EAAE,CAAA;IAEvD,MAAM,QAAQ,GAAG,qBAAqB,CAAC,mBAAmB,CAAC,aAAa,EAAE,EAAE,CAAC,CAAA;IAC7E,MAAM,eAAe,GAAG,MAAM,CAAC,KAAK,CAAC,UAAU,GAAG,CAAC,CAAC,CAAA;IACpD,MAAM,2BAA2B,GAAG,MAAM,CAAC,MAAM,CAAC,CAAC,QAAQ,EAAE,eAAe,CAAC,CAAC,CAAA;IAC9E,MAAM,iBAAiB,GAAG,2BAA2B,CAAC,MAAM,CAAA;IAE5D,MAAM,YAAY,GAAG,UAAU,GAAG,GAAG,CAAA;IAErC,IAAI,kBAAkB,GAAG,EAAE,CAAA;IAE3B,KAAK,IAAI,UAAU,GAAG,CAAC,EAAE,UAAU,GAAG,iBAAiB,EAAE,UAAU,IAAI,YAAY,EAAE;QACpF,MAAM,SAAS,GAAG,IAAI,CAAC,GAAG,CAAC,YAAY,EAAE,iBAAiB,GAAG,UAAU,CAAC,CAAA;QAExE,MAAM,KAAK,GAAG,2BAA2B,CAAC,QAAQ,CAAC,UAAU,EAAE,UAAU,GAAG,SAAS,CAAC,CAAA;QAEtF,MAAM,WAAW,GAAG,MAAM,UAAU,CAAC,mBAAmB,CAAC,KAAK,CAAC,CAAA;QAE/D,IAAI,OAAO,EAAE;YACZ,MAAM,iBAAiB,GAAG,UAAU,CAAC,aAAa,EAAE,CAAC,OAAO,CAAA;YAE5D,IAAI,iBAAiB,IAAI,kBAAkB,EAAE;gBAC5C,+BAA+B;gBAC/B,gBAAgB;gBAEhB,kBAAkB,GAAG,iBAAiB,CAAA;aACtC;SACD;KACD;IAED,uFAAuF;IAEvF,MAAM,MAAM,GAAG,UAAU,CAAC,WAAW,EAAE,CAAA;IAEvC,UAAU,CAAC,KAAK,EAAE,CAAA;IAClB,UAAU,CAAC,IAAI,EAAE,CAAA;IACjB,KAAK,CAAC,IAAI,EAAE,CAAA;IAEZ,MAAM,UAAU,GAAW,MAAM,CAAC,IAAI,CAAA;IACtC,MAAM,MAAM,GAA2B,MAAM,CAAC,MAAM,CAAA;IAEpD,IAAI,MAAM,CAAC,MAAM,IAAI,CAAC,EAAE;QACvB,OAAO,EAAE,UAAU,EAAE,QAAQ,EAAE,EAAE,EAAE,CAAA;KACnC;IAED,MAAM,QAAQ,GAAa,EAAE,CAAA;IAE7B,KAAK,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,GAAG,MAAM,CAAC,MAAM,EAAE,CAAC,EAAE,EAAE;QACvC,MAAM,KAAK,GAAG,MAAM,CAAC,CAAC,CAAC,CAAA;QAEvB,MAAM,SAAS,GAAG,KAAK,CAAC,IAAI,CAAA;QAC5B,MAAM,UAAU,GAAG,KAAK,CAAC,KAAK,CAAA;QAC9B,MAAM,QAAQ,GAAG,KAAK,CAAC,GAAG,CAAA;QAC1B,MAAM,eAAe,GAAG,KAAK,CAAC,IAAI,CAAA;QAElC,QAAQ,CAAC,IAAI,CAAC;YACb,IAAI,EAAE,MAAM;YACZ,IAAI,EAAE,SAAS;YACf,SAAS,EAAE,UAAU;YACrB,OAAO,EAAE,QAAQ;YACjB,UAAU,EAAE,eAAe;SAC3B,CAAC,CAAA;KACF;IAED,wEAAwE;IAExE,MAAM,CAAC,GAAG,EAAE,CAAA;IAEZ,OAAO,EAAE,UAAU,EAAE,QAAQ,EAAE,CAAA;AAChC,CAAC"}
|