Hi, I’m developing a speech recognition service with galaxy buds series(bluetooth earset).
There seems to be a mismatch between ASR train data and speech signal of galaxy buds’ microphone.
I think that galaxy buds’ microphone has applied signal processing modules to improve telecommunication experience, but it degraded ASR performance.
So, I wonder how Samsung is dealing with this problem when using galaxy buds, Whether buds data logs are collected or specific signal processing is applied.