• Pytorh GPU implementation of CNN for Motor-Tic Detection in Movement Disorders
• Eye Tics vs Normal Classification

• All source code is in src.
• Train using the CNN.py file.
• Get best model using the TestCNN.py file.

• Code is the documentation of itself.

• Use python3 CNN.py to generate a confusion matrix.
• A summary of the pipeline is given in report.pdf.

The pipeline is demonstrated below.

• Training Curves.

Losses	Accuracies

• Classification Confusion Matrix.

Absolute Values	Percentages

• apost035@umn.edu, trs.apostolou@gmail.com