ImageNet Large Scale Visual Recognition Challenge

This competition would play an out-sized role in this story, as the little-known event was responsible for the decade’s most important machine vision breakthrough in 2012, when deep convolutional neural networks, a technique first built in the 1980s, won the competition outright. ImageNet was also responsible for Neural Style Transfers in 2015. The ImageNet Large Scale Visual Recognition Challenge (ILSVRC) began in 2010 and lasted annually until 2017. The competition evaluated algorithms for large scale object detection and image classification. It utilized a subset of the vast ImageNet dataset—mentioned in the previous as an “All-Time Top Moment” and critical in the development of AI art—which at this point included over 10 million labeled images across over 10,000 categories. The competition aimed to benchmark advancements in computer vision with participants' algorithms identifying objects in unlabeled test images, producing a list of probable categories. This challenge drove significant progress in image recognition technologies, fostering innovation and providing a platform for researchers to compare their work against state-of-the-art methods in object detection and image annotation.