ML Club Video: Convolutional Neural Networks!

When you look at a face, what do you see? You see eyes, ears, a nose, and other facial features. When a computer normally looks at a photo of a face, it sees pixels – individual blocks of color arranged in an organized array. But this method isn’t as effective as our “human” method is – even from a deep-learning perspective. Therefore, we use Convolutional Neural Networks (CNNs) – which use sliding windows – to pick up on features rather than pixels, and to find relationships between different parts of the image to make their prediction.

These CNNs are used everywhere – they are the very foundation of today’s computer vision models, being used in applications from facial recognition to self-driving cars!

image from <https://learnopencv.com/understanding-convolutional-neural-networks-cnn/

There is an amazing website that explains all about CNNs – here it is!

https://poloclub.github.io/cnn-explainer/

More interested? Watch the ML Club video below!