Closed jonwalch closed 4 years ago
http://cs231n.stanford.edu/reports/2016/pdfs/273_Report.pdf
They specifically cut out the scoreboard for this paper so the model couldn't learn using it and it has to use just the video. Academia...
Papers that cited this one: https://scholar.google.com/scholar?cites=15590873559941068473&as_sdt=2005&sciodt=0,5&hl=en
Going to prototype a ternary model with three buckers {no_win, ct_win, t_win}. I need to find/generate a large amount of images of counter strike. I need screens of anything that isnt a round end, cts winning, and ts winning. I need about 1/3 of each type of image.
https://pytorch.org/tutorials/beginner/data_loading_tutorial.html
Collect lots of image of: no win CT win Ts win
should crop them to the parts that are relevant when doing inference can crop input image stream proportionally
Closing in favor of https://github.com/jonwalch/whiplash/issues/462
Goal: build an ML model that can determine whether Ts or CTs win the round