We can run the script as python fvi_computation.py.

We install dependencies as "pip install numpy opencv-python"

The second script is for first selecting the top k frames from each cluster and then subsequently applying FVI.