From the course: Neural Networks and Convolutional Neural Networks Essential Training
Unlock the full course today
Join today to access over 22,600 courses taught by industry experts or purchase this course individually.
Zero padding and pooling
From the course: Neural Networks and Convolutional Neural Networks Essential Training
Zero padding and pooling
- [Instructor] So let's talk a little bit about Zero Padding. During the forward pass, we slide, or convolve, each filter across the width and height of the input volume, and compute dot products between the entries of the filter and the input at any position. So, for example, we'll take the four from the top corner of our three by three filter, and multiply that by zero in the image. And then move your cross, we will multiply the zero on the filter with the corresponding location on the image which is also a zero, and so on. And we will do that for all of the nine entries in the kernel. We then sum up all of the product or multiplication calculations, and this known as the dot product. As we slide the filter over the width and height of the input volume, we will produce a two dimensional activation map that gives the responses of that filter at every spacial position. Here's an example of what convolving looks like for a seven-by-seven image with a three-by-three filter. We have used…
Practice while you learn with exercise files
Download the files the instructor uses to teach the course. Follow along and learn by watching, listening and practicing.