My goal of studying deep learning is developing some fun computer vision(object detection, recognition, background, foreground separation etc) tools, as far as I know cnn is somehow like the defacto standard for many computer vision tasks but not rnn, dose it matter if I skip the part of rnn? Would it stop me from understanding cnn?
Well, people are combining both rnns and cnns to do insane stuff . Attentional models that jeremy mentions in part-1(which will be covered in part-2), video understanding are just few examples of rnns and cnns combined.
Thanks, looks like I should not skip rnn.