Clarification on CNN

mahmoud_ahc · September 6, 2021, 12:11am

Quoting this from this resource, would you please explain it to me?

Real-world example. The Krizhevsky et al. architecture that won the ImageNet challenge in 2012 accepted images of size [227x227x3]. On the first Convolutional Layer, it used neurons with receptive field size F=11, stride S=4 and no zero padding P=0. Since (227 - 11)/4 + 1 = 55, and since the Conv layer had a depth of K=96, the Conv layer output volume had size [55x55x96]. Each of the 555596 neurons in this volume was connected to a region of size [11x11x3] in the input volume. Moreover, all 96 neurons in each depth column are connected to the same [11x11x3] region of the input, but of course with different weights.

edwardyu · September 13, 2021, 9:47am

Hi @mahmoud_ahc ,

You can use Week1: Simple Convolutional Network Example as an analogy, just like below (red number with parentheses is parameters in the paper).

Topic		Replies	Views
Simple Convolutional Network Example Convolutional Neural Networks	2	515	May 16, 2022
C4W1 Convolutional Neural Network Example - Typo/Error? Convolutional Neural Networks	4	593	April 21, 2022
Lecture Question - Convolutional Implementation of Sliding Windows Convolutional Neural Networks	3	545	June 7, 2022
Course 4: Week 1: CNN Example: CONV1 layer Activation Size Convolutional Neural Networks	1	494	September 11, 2022
Question about channel dimension matching Convolutional Neural Networks	4	381	September 4, 2023

Clarification on CNN

Related topics