Kaljurand / dictate.js

A small Javascript library for browser-based real-time speech recognition, which uses Recorderjs for audio capture, and a WebSocket connection to the Kaldi GStreamer server for speech recognition.
https://kaljurand.github.io/dictate.js/
BSD 3-Clause "New" or "Revised" License
216 stars 62 forks source link

It is not showing my kaldi words #17

Closed lio14cha closed 5 years ago

lio14cha commented 6 years ago

I tried the demo.html. But it is not following my kaldi words. How to connect it to my own ASR. how to configure this demo.html. It is showing like msg: 10: 1005//true msg: END OF SESSION: msg: 8: {"status": 0, "adaptation_state": {"type": "string+gzip+base64", "id": "71d55971-c898-402f-b0f6-5dda543b487c", "value": "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", "time": "2018-08-01T08:13:38"}, "id": "71d55971-c898-402f-b0f6-5dda543b487c"} msg: 7: Send tag: EOS msg: 5: Send: blob: audio/x-raw, 2972 msg: END OF SPEECH: msg: 11: Stopped recording msg: 5: Send: blob: audio/x-raw, 8916 msg: 5: Send: blob: audio/x-raw, 8918 msg: 5: Send: blob: audio/x-raw, 5944 msg: 5: Send: blob: audio/x-raw, 8916 msg: 5: Send: blob: audio/x-raw, 8916 msg: 5: Send: blob: audio/x-raw, 5944 msg: 9: [object Event] msg: READY FOR SPEECH: msg: 3: Recorder initialized msg: 2: Media stream created msg: 1: Waiting for approval to access your microphone ... after speaking.No output text is presenting on the textarea. Some times some unusual words are presenting. Thank you.

Kaljurand commented 6 years ago

You seem to be connected to the Kaldi server because adaptation_state is returned. More details are needed to debug this further. Does the server work, e.g. can you use it with the official client Python script? Does recording in the browser work, e.g. can you use speech recording/recognition with some other webservices?