@barloccia co-author
Raspberry Pi based project, was born as a joke and grown up quickly and returned unexpected good results.
- Find a structure and find out how manage engines with Raspberry:
main.py
is the routine. - Develop and improve offline speech-recognition, Snowboy used:
rpi/
andosx/
contains the framework, - Develop a user interface for train models and manage the car:
app/
Snowboy (https://snowboy.kitt.ai) is an highly customizable hotword detection engine that is embedded real-time and is always listening (even when offline) compatible with Raspberry Pi. For the purpose of recognize key words, snowboy use a trained models that can be improved by a community based strategy.
This interface was created with the purpose to manage raspberry by remote (see app/synchronizer/
and app/ssh_conn
)
and train snowboy models (see app/train_routine
). The rest of code is stuff that put other stuff togheter.
To achieve above interface PyQt5
were used, follows versions of requirements used:
Requirement | Version Used |
---|---|
Python | 3.6 |
PyQt | 5.10 |
pysftp | 0.2.9 |
PyAudio | 0.2.11 |
QDarkStyle | 2.3.1 |