Mp4grep – search in transcriptions of audio/video files
github.com
github.com
Most of offline the Speech to Text tools I know have a bad model that will have a hard time at understanding anything that is not English :(
Mozilla DeepSpeech seems to have well trained models, from what I remember.