FaceTime

Build

cd env/

Start the virtual environment: source bin/activate

If any dependancies are not installed run: pip install -r requirements.txt

if you add more dependancies run: pip freeze > requirements.txt

Once you're finished working with the virtual env run: deactivate

Run

1. In facerecon-system.py change the frame-rate to the desired fps to pull frames (default every 0.5s)
2. In run.sh enter the desired time-stamps of the video you would like to process
3. Modify run.sh with the video you would like to process
4. Make a new folder called crop New Folder > crop
5. ./run.sh to run the cropper and wait as it fills your new crop folder

---

baseline.ipynb contains the code to run the baseline model

hypertuning.ipynb contains the code to run the binary models on particular actors

Set desired=# with the following lookup to train on that particular actor:

'Leonard': 1, 'Penny': 3, 'Raj': 4, 'Sheldon': 5, 'Wolowitz': 6

onscreen.ipynb contains the code to find the final onscreen time for each actor on an episode