Formula for Receptive Field

#1

For an input image, the receptive field is 3x3 for next conv layer, and for next its 5x5. and for third conv layer receptive field is 7x7. this is the case when kernel_size is 3x3 stride =1, padding =0
can you give a general formula to find receptive field for a particular conv layer, having stride s, and padding p.
Thank you

#3

No, this formula will give the size of convovled layer,
but i am asking, if we are applying 5 convolution layer, kernel_size = (4,4) each with stride 2 and padding (2,2) then what would be the receptive field for 1 pixel after all these layers. means how much part of the original image, that one pixel would be able to see.

#4

Oh, I misunderstood that question :sweat_smile:
This link here has explained the same very well.

Happy Learning :slight_smile:

#5

okay, I will read the blog. :slight_smile:
Thanks for the link

#6

I hope I’ve cleared your doubt. I ask you to please rate your experience here
Your feedback is very important. It helps us improve our platform and hence provide you
the learning experience you deserve.

On the off chance, you still have some questions or not find the answers satisfactory, you may reopen
the doubt.