search

aadiltajani / CodeSten

CodeSten is Project1 for Group19 of CSC510 Software Engineering
GNU General Public License v3.0
0 stars 0 forks source link

SR Recognizer audio length limit encountered #3

Closed aadiltajani closed 2 years ago

aadiltajani commented 2 years ago

Google's Speech Recognizer is an open source package for speech recognitionn and allows only 30 seconds of audio data to be worked upon in a single request. So for longer audio files, it won't allow to work on it. Need to find a better open source package or have to find a solution for it

Dhruvish-Patel commented 2 years ago

Using another open source package may affect our accuracy. Alternative way to address this issue can be to break the audio data into smaller chunks and iteratively combining the output of the smaller chunks.

aadiltajani commented 2 years ago

Yepp good idea. If anyone can come up with a good splitting method lets implement it

Dhruvish-Patel commented 2 years ago

Added a trim_audio_file method in readaudio.py file to trim the audio into smaller chunks.

aadiltajani commented 2 years ago

Checked it... its working