What are the ways to improve the speed of continuously recognizing multiple audio files?

[Morgen92125]

When I used the CPU for speech recognition, I found that there was a delay of over 2 seconds for the shortest speech file (about 5 words), and the average recognition time for each sentence in a long file was much less. I guess it was caused by loading and starting the model. Is it possible to recognize multiple files in a row without spending over 2 seconds each time loading and starting the model?

[trungkienbkhn]

@Morgen92125 , hello. You can use the multithread in Python to execute each audio file