• https://www.liip.ch/en/blog/speech-recognition-with-wit-ai
• https://wit.ai

• Install pip, python and pip libraries

sh install.sh

• queued_from_file (best mode):

Record speaker audio every 10 seconds on a file with timestamp name inside ./sounds folder

sh record-from-speaker.sh

Watch folder ./sounds, send each new file to WIT.AI and remove them

sh watch-records-and-recognize.sh

• live_from_file: Record speaker audio every 10 seconds on same record.wav file

sh record-from-speaker.sh

Send same file to WIT.AI every 10 seconds

python main.py

• live_from_mic: Record microphone every 10 seconds on record.wav file and send to WIT.AI

python main.py