Convolution Defined — Introduction to Convolutional Neural Networks | by Egor Howell | Dec, 2023
The basic constructing block of CNNs
My latest articles have been a collection on neural networks the place we go from the easy perceptron to sophisticated architectures and the way to cope with common problems in deep learning. In case you are , be happy to test the collection right here:
Neural Networks
One thrilling space neural networks have made vital strides in is computer vision. Assume AI for self-driving vehicles and face recognition!
Nonetheless, the common totally linked neural community that most individuals find out about will not be appropriate for a lot of real-life picture recognition duties. It really works on the well-known MNIST dataset, but it surely has small photographs of 28×28 pixels.
Excessive-definition (HD) photographs have 1280×720 pixels. That’s roughly 1,000,000 pixels, which might imply 1,000,000 neurons within the enter layer. To not point out the thousands and thousands of weights required for the hidden layers, rendering common neural networks unsuitable because of the dimensional complexity.
So, what will we do?
Convolutional Neural Networks!
Convolutional neural networks (CNN) are the gold customary for almost all of pc imaginative and prescient duties at the moment. As an alternative of totally linked layers, they’ve partially linked layers and share their weights, decreasing the complexity of the mannequin.
For example, for every neuron in a totally linked neural community layer, we’d require 10,000 weight of a picture of 100×100 pixels. Nonetheless, a CNN can have solely 25 neurons to course of the identical picture.
On this article, we’re going to dive into the basic constructing block behind CNNs, convolution.
Like many issues in machine studying, CNNs are impressed by nature. Pc scientists checked out how the visual cortex in the mind works and utilized the same idea to neural networks.
