At the NIPS academic conference held in 2012, Alex Krizhevsky and his collaborators, one of whom was the neural network pioneer, Geoffrey Hinton, presented a record breaking result at the ImageNet Large-Scale Visual Recognition Competition (ILSVRC). Research teams competed in various image recognition tasks that used the ImageNet dataset. Krizhevsky's results on the image classification task were 10.8% better than the state of the art. He had used GPUs for the first time to train a CNN with many layers. This network structure would popularly be called AlexNet later. The design of such a deep neural network with a large number of layers is the reason why this field came to be called deep learning. Krizhevsky shared the entire source code of his network, now called cuda-convnet, along with its highly GPU-optimized training code.
Soon...
