skit-ai / kaldi-serve

Server framework for Kaldi ASR Toolkit
Apache License 2.0
97 stars 24 forks source link

Speech end-pointing #37

Open greed2411 opened 2 years ago

greed2411 commented 2 years ago

To be updated.

pskrunner14 commented 2 years ago

Task 1

Implement Speech Endpointing feature in the Decoder class of kaldi-serve core C++ library. Speech Endpointing is the process of deciding when to stop decoding based on trailing silence in the audio.

Relevant links:

  1. Online Decoding TCP Server Binary
  2. Online Endpoint - Source

Task 2

Add a boolean flag in kaldi-serve proto file that toggles the Speech Endpointing feature (Task 1) in all the RPC methods implemented in the kaldi-serve gRPC server application.