I did a project of object detection in which there was a single object in the whole image. My CNN network was taking an image and was outputting two things one is the class or category of the object (e.g dog, cats etc. ) and the second was the bounding box coordinates. It was very simple to implement coz all the images had only one object and thus only set of bounding boxes.
So I thought of implementing multiple object detection where all the images can have one or more than one object. So the issue I was facing was when I was making matrices of classes let's say I have two categories so my class matrix would be [ [0], [0,1], [ 1,1] ….] 0 for cats and 1 for dogs but the matrices are unequal coz there can be one cat in an image and in another it can be two or 4 etc. as the number of objects in the images are not fixed . Same goes for the bounding boxes also.
So how should I train my network ?
Best Answer
Finally I got my answer...the concept of multiple object detection using anchor boxes is very well explained here d2l.ai/chapter_computer-vision/anchor.html
It is a very amazing book Dive into Deep Learning released by Amazon.