Deep Convolutional Neural Networks Unleashed

Arguably the biggest development in machine learning in decades, deep convolutional neural networks (CNNs or “ConvNets”) revolutionized machine vision. Their power was deployed for the first time relating to creative expression when the SuperVision algorithm from the University of Toronto used CNNs to win the ImageNet Large-Scale Visual Recognition Challenge. Critically, ConvNets were the driving force behind the decade’s upcoming AI developments This paradigm-shifting technique enabled computers to recognize objects with near-human accuracy by using multiple layers of small neuron collections to analyze images. Each layer overlaps and refines the image representation, allowing the system to learn intricate details. Originally conceived of theoretically in the 1980s, CNN research for the thirty years was mostly quiet due to computing restraints. They only gained prominence as computing power advanced, enabling the analysis of the large datasets required. This moments serves as the main milestone in AI art developments in the decade along with GANs in 2014 and DeepDream in 2015. It also highlights the importance of 2009’s ImageNet—covered in the previous chapter—which had a lasting impact on the space due to its Large-Scale Visual Recognition Challenge that began in 2010.