OWLgameplay

This is a silly project that started with me and my brothers watching overwatch league matches and complaining about how terrible the casters are at filling time. I was looking for a new data science personal project and figured why not try and solve our problem with machine learning. In this notebook, I've written a convolutiuon neurel network to identify gameplay and skip all other parts of the match. It actually works pretty well. here is a link to me rambling about my process for the first version I put up on github

https://www.ndled.us/stats/2021/08/06/CV-OWL.html

Things I read while doing this

1. http://zulko.github.io/blog/2014/07/04/automatic-soccer-highlights-compilations-with-python/

2. https://www.udemy.com/course/machinelearning/learn/lecture/6761136

3. https://www.toptal.com/machine-learning/machine-learning-video-analysis

4. https://www.analyticsvidhya.com/blog/2018/09/deep-learning-video-classification-python/

5. https://machinelearningmastery.com/how-to-visualize-filters-and-feature-maps-in-convolutional-neural-networks/

You will find the jupyter notebook I used while developing this as well as a command line tool for automation purposes.