Sliding Window v/s Convolution

[Rakhshanda_Mujib]

When creating a post, please add:

• Week #3
• Link to the classroom item you are referring to: https://www.coursera.org/learn/convolutional-neural-networks/lecture/6UnU4/convolutional-implementation-of-sliding-windows
• Description: How is the sliding window conceptually different from convolution? In case we’re sliding, we are feeding individual crops to the ConvNet but aren’t we doing the same when it comes to convolving the whole image and passing it through the ConvNet? I understand there is some level of parallellism but isn’t sliding and convolving essentially the same thing?

[TMosh]

Buy itself, a sliding window is just a data clipper.
A convolution also uses sliding, but it goes further and also transforms the data.

[rmwkwok]

As the title suggested - “Convolutional Implementation of Sliding Windows”, they are achieving the same goal in two different ways!

Cheers,
Raymond

[locquan.ai]

I have the same question. Rewording for clarity, the question would be “Is the convolution implementation of sliding window algorithm the same as a normal ConvNet?”.

To my knowledge and research, they seem to be the same. Normal ConvNets (CNNs) have implicit sliding window mechanism under (by means of convolution operations). I’m also a learner and no expert, so I could be wrong though.