search

alumae / kaldi-gstreamer-server

Real-time full-duplex speech recognition server, based on the Kaldi toolkit and the GStreamer framwork.
BSD 2-Clause "Simplified" License
1.07k stars 342 forks source link

chain models poor accuracy and big latency #164

Open lucgeo opened 5 years ago

lucgeo commented 5 years ago

Hello,

Using the chain models and a yaml config file similar to sample_english_nnet2.yaml (the only major difference is "nnet-mode:3"), the transcriber is working with poor accuracy and big latency. NNET2 models together with a yaml file with the same structure are performing well. Could you provide a correct model of yaml file for chain models? Which may be the cause of this issue?

Thank you!

boleamol commented 5 years ago

Hi @lucgeo , I am also facing same problem. Have you solved it? if so then please send me solution.

lucgeo commented 5 years ago

Hi,

Unfortunately no, I still have that problem.

În mie., 29 mai 2019 la 12:14, Amol Bole notifications@github.com a scris:

Hi @lucgeo https://github.com/lucgeo , I am also facing same problem. Have you solved it? if so then please send me solution.

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/alumae/kaldi-gstreamer-server/issues/164?email_source=notifications&email_token=ABUASYKBEW5VPWEY73YPFELPXZQQTA5CNFSM4GKGHRGKYY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGODWO7Z6I#issuecomment-496893177, or mute the thread https://github.com/notifications/unsubscribe-auth/ABUASYJ4ZO7BUTGANJ2W5ETPXZQQTANCNFSM4GKGHRGA .

sirifarif commented 4 years ago

Any update on this issues. specially dealing with the latency of the recognition.