Bug fix for audio files which have invalid (or empty) transcription preceding valid segments of speech.
This happened in cases when a file has some background noise or dialogue, followed by some static or silence, followed by the actual audio/dialogue we want to transcribe.
CRIS was returning a transcription for the background audio which had <5% confidence resulting in an empty/blank transcription.
Bug fix for audio files which have invalid (or empty) transcription preceding valid segments of speech.
This happened in cases when a file has some background noise or dialogue, followed by some static or silence, followed by the actual audio/dialogue we want to transcribe. CRIS was returning a transcription for the background audio which had <5% confidence resulting in an empty/blank transcription.