OhSnap

Predicts who snapped in real-time using audio.

How to use

1. Unzip the given dataset snaps.zip.
2. Train model using train.py, adjust threshold and parameters.
3. Run a live demo using demo.py
4. Now you can create your own dataset for recognition! Change the paths in train.py

Built With

• cepstrum
• cool-features
• machine-learning
• pc-microphone
• python
• random-forests
• scikit-learn
• tkinter