Filters inside YOLOv5 architecture

[shubham-scisar]

Do the layers i.e. convolution layers use intrinsically any filters such as Edge detection,blur inside the architecture ?

[glenn-jocher]

@shubhamsharma1609 no. Deep learning models learn their own feature extraction convolution kernels, that's the whole point of deep learning in vision AI.

[shubham-scisar]

@glenn-jocher Apologies if my question was trivial. I agree with your explanation,but I was just trying to explore the architecture and object detection through yolov5 (with internal working) and could not find any documentation related to the same.