speech-rec.sh is a bash script that sends a flac audio file to Google for speech recognition and prints the N-best returned hypotheses. It uses the (until now) freely available, though up to 50 requests a day limited, Google ASR API version 2.

More interestingly, if you have sox (or parecord and timeout) installed on your machine, you can use the script to interactively record and utterance and get its transcription.

To install sox type:

sudo apt-get install sox

For more details about Google's ASR service check out this article.

./speech-rec.sh -i record.flac --rate 16000

Use option -l to set language:

French

./speech-rec.sh -i record.flac --rate 16000 --language fr_FR

Spanish

./speech-rec.sh -i record.flac -r 16000 -l es_ES

German

./speech-rec.sh -i record.flac -r 16000 -l de_DE

This will give you 3 seconds to talk:

./speech-rec.sh

Use option -d to set duration:

./speech-rec.sh -d 7

The script writes audio data into an audio file named record_DATE_TIME.flac where DATE_TIME is the output of:

date "+%Y%b%d_%H-%M-%S"

To replay your utterance with sox type:

play record.flac

or with paplay:

paplay record.flac

The key delivered with this script is to be used for test purposes only. It may be disabled by Google anytime. You should generate and use your own API key (follow this link for key generation instructions: http://www.chromium.org/developers/how-tos/api-keys)

for audio data acquistion, or

if you don't have sox.

to send data to server.

this repository offers a Python ASR API for many vendors.

Amine SEHILI ([email protected])

Copyright (C) Amine SEHILI 2015-2016.

This program is available under the GNU GENERAL PUBLIC LICENSE Version 3.

Source code on GitHub: https://github.com/amsehili/gspeech-rec.