I am trying to implement sox into a speech recognition program. However it tooks nearly 10 seconds for sox to trim words in very simple sentence with 5-7 words (file size is also between 200-400kb) it is unacceptaable. What can I do, or which other options can be used for the same aim. by the way the sox parameters I am using are
sox -c 1 -r 16000 input.wav output.wav silence 1 0.5 1% 1 0.1 1% : newfile : restart