Lec 17 : Introduction to Convolution Neural Networks
This video answers following questions with toy examples and intuitive explanations:
1- Why we need CNN when we have Fully connected Neural Nets?
2- Why use Convolution and not any other operation?
3- What is convolution? Solved toy example of Convolution
4- How convolution reduces dimensions of image?
5- How padding is used to retain original dimensions of image after convolution?
6- What are the benefits of padding?
7- How can we calculate required padding size for image?
