Crash Course: Neural Networks Part 6 -- Convolutional Neural Networks

Computers may be smarter at some tasks than humans are, but image classification is not the case. For us, it's extremely easy to detect a plane, or a cat, or a dog. For a computer, which works only with numbers, that task is extremely difficult. For computers to be able to detect images, we had to take some inspiration from nature, from how neural cells and our eyes process images. That's how Convolutional Neural Networks were born, which are now applied to a lot more tasks than only visual recognition.