I am in a very similar situation regarding both
- my telephony expertise and
- that I want to use Asterisk solely for sending/receiving the caller’s audio to/from an external server as a stream (realtime audio passing i.e. no recordings) - meaning the external server could basically be replaced by a human on another phone.
From what I’ve read so far, I also get the impression that a lot of approaches out there aim to solve either speech-to-text (STT) or text-to-speech (TTS) but not both (STS). For me, it’s also unclear which interface should be used for what purpose…
Which of the following approaches is the easiest/best for Asterisk for STS?
- WebRTC
- AEAP
- ARI
- UniMRCP
Related threads:
- Transcribe Audio to File - Realtime
- from 2023, suggests ARI with external media