bsenftner
commented
2 weeks ago This repo was actually a homework assignment from an OpenCV course I took at openCV.org. That specific class is no longer offered, but this one covers the same information: https://opencv.org/university/computer-vision-and-deep-learning-applications/
For what it is worth, my programming background includes extensive 3D computer graphics and video programming. I worked at Philips Media while mpeg was being developed and participated in those early drafts of digital video. After Philips Media, I was at Electronic Arts, where I wrote the video subsystem for the 3D0 game console, and then I did the same for the original Sony PlayStation. Then about 6 years after that, I was part of the development of live video over the Internet, where I wrote one of the original time code accurate multiple video resolution streaming codecs.
Skipping over two entire careers, 15 years after that Live Video over the Internet work I was principal software scientist for one of the top 5 global facial recognition providers. There I wrote a variation of this: https://github.com/bsenftner/ffvideo Which is an optimized ffmpeg player, which can be coupled with the logic in this repo and get pretty much a 7-years-ago state of the art (then) facial recognition system (minus the petabyte of facial images for training.)
I was reading about ffmpeg when i stumbled upon this, just wanted to say thanks for publishing this repo. It helped me understand a lot of new concepts, i took the python notebook and applied it in CPP and it worked!
Could you please provide me with additional reads or ideas to deepen my understanding of this subject, if possible?
Thanks in advance!