Entrepreneur, Software Developer, Machine Learning and Computer Vision Researcher. Contact me: firstname.lastname@example.org
Stream and video is the same thing, just frames. You can use any video or stream, but need to change markup of detection zones and params of video. Its not a production ready app, its just a simple example
Also i dont include python courses, and basic math) Cause it’s very basic, if you dont know basic stuff then ML is not what you need right now.
try on free GPU here https://colab.research.google.com
Dont use MacOS for ML tasks, its really bad
What do you mean manually? sending data with hardware only?
also i recommend you to look at this paper https://groups.csail.mit.edu/graphics/hdrnet/
What about bad examples from this work, cause handpicked samples not the best way to show results.
i dont have mi3 so i cant test it, but if you have mi3 you can add lib for mi3 based on current article.
yes of course)
Model which getting more data will have more probability to give better result