cmu-sphinx-decoder

CMU Sphinx decoder

pocketsphinx_parallel_decoder.py

Decode audio files in parallel. This command load N decoders that run in parallel, where each one decode audio files sequentially.

usage: pocketsphinx_parallel_decoder.py [-h] -i INDIR -c CONF [-l] [-j JOBS] -o OUT

The following arguments are required: -i/--indir, -c/--conf, -o/--out

Arguments:

-i/--indir

The input audio directory. Accepted formats: any format that ffmpeg supports. If the audio files are not in wave format, the code will convert it on the fly (during running time) to /tmp, then they will be deleted after decoding.

-c/--conf

This argument is for the configuration file (ini). You can specify the path for HMM and LM models as well as the dictionary.

-o/--out

specify the output directory. A suffix with process id/part id will be added automatically to the file name.

-j/--jobs

specify the number of jobs (parallel decoders) if j > number of CPUs/Cores then an error message with appear. j should be <= number of CPUs/Cores. This argument provides the user with the flexibility to control the load on the machine.

-l/--log

enable logging (for debugging purposes)

example command

python cmu-sphinx-decoder/pocketsphinx_parallel_decoder.py -j 8 -c cmu-sphinx-decoder/conf/config_ar.ini -i /storage/recordings/2713/2018/08/16  -o decoder_out/
nohup python cmu-sphinx-decoder/pocketsphinx_parallel_decoder.py -j 8 -c cmu-sphinx-decoder/conf/config_ar.ini -i /storage/recordings/2713/2018/08/16 -o decoder_out/ &    
python cmu-sphinx-decoder/pocketsphinx_sequential_decoder.py -c cmu-sphinx-decoder/conf/config_ar.ini -i /storage/recordings/2713/2018/08/16  -o decoder_out/

Error codes:

-1 : config file does't exist

-2 : HMM file does't exist

-3 : language model file does't exist

-4 : dictionary file does't exist