For the past few days, I have been learning a little bit more about R-CNNs, Fast-RCNNs, and Faster R-CNN. The three techniques are really interesting the more I learn about their inner-working.
The most surprising thing is just how good they are at creating annotations for unlabelled data. It does not seem like the images being fed into those networks have any annotations on them, the model just kinda…figures it out by itself, which I think is pretty dope.
ML Devlog, Episode 2: R-CNN, part 1
This post is licensed under
CC BY 4.0
by the author.