CNN architecture that returns a number of objects
I am looking for an architecture of neural network that can return a number of objects in an image. It can be not only vanilla CNN but any other kind of NN. I already found articles about crowd counting, I'm looking for articles about different types of objects. Thanks.