Three-dimensional convolutions
Now that we have a good idea of how two-dimensional convolutions work, it is time to move on to three-dimensional convolutions. But wait—didn't we just learn about three-dimensional convolutions? Kind of, but not really because, if you remember, they had the same depth as the volume we were convolving over and moved the same as the two-dimensional convolutions did—along the height and the width of the image.
Three-dimensional convolutions work a bit differently in that they convolve over the depth as well as the height and the width. This tells us that the depth of the kernel is smaller than the depth of the volume we want to convolve over, and at each step, it performs element-wise multiplication and addition, ...
Become an O’Reilly member and get unlimited access to this title plus top books and audiobooks from O’Reilly and nearly 200 top publishers, thousands of courses curated by job role, 150+ live events each month,
and much more.
Read now
Unlock full access