I am learning cs231n Convolutional Neural Networks for Visual Recognition. The lecture notes introduce the concepts of width, height, depth. For example,

In CIFAR-10, images are only of size 32x32x3 (32 wide, 32 high, 3 color channels)

However, in another example,

a volume of size [55x55x96] has 96 depth slices, each of size [55x55]

What does width 96 means? Does it mean 96 color channels? Why can we have more than 3 color channels?

  • 1
    $\begingroup$ No, 96 is the number of neurons that are looking 55x55 region in the depth of the input image and generate a feature map. You can say that you have 96 filters applied across the depth of the image and learning some feature maps. $\endgroup$ – Nomiluks Apr 7 '16 at 21:32

It means that the number of filters (a.k.a. kernels, or feature detector) in the previous convolutional layer is 96. You may want to watch the video of the lecture, and in particular this slide, which mentions that a filter is applied to the full depth of your previous layer:

enter image description here


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.