Application to control linux OS using Deep Learning
Achieved an accuracy of 80% using both CNN and LSTM for 30 classes of sounds.
- main.py
- Audio Handler
- Text Parser
- Action Executor
- DNN Model
- Model Preprocessor
- Dataset Handler
Command Mode is activated as soon as we say "Marvin". To get out of this command mode we can say "Marvin Stop".
| Action | Command |
|---|---|
| Sleep | tree stop |
| Volume Up | shiela up |
| Mute | shiela down |
| Scroll Up | bird up |
| Scroll Down | bird down |
| Press "Enter" | yes |
| Press "Escape" | no |
| Take Screenshot | wow |