I tried Google speech-to-text for few hindi and telugu phone calls. It's missing out few sentences in the transcript. In the below image, we can see that 01:52 - 02:01
and 01:52 - 02:01
etc parts are missed in the transcript. I am facing same issue irrespective of any format audio (mp3, wav, stereo, mono).
Looks like it is giving only the parts where confidence>0.8 . But I didn't set that confidence threshold anywhere while in the console config while generating. So, is there any way where I can get the transcript without missing parts in it?