ImageNet Classification with Deep Convolutional…

AlexNet asks how a deep image classifier can learn from 1.2 million examples without taking impossibly long to optimize or simply memorizing the training set. Read it to see how ReLU activations, a selectively connected two-GPU network, image augmentation, and dropout made an eight-layer convolutional model practical at ImageNet scale.

Reading focus: Why ReLU reached a fixed training-error threshold much sooner than tanh in the paper's controlled comparison, without implying one universal speedup. How two 3 GB GPUs divided the feature maps, exchanged information at selected layers, and converged into one 1,000-class classifier. How random crops, reflections, color perturbations, dropout, test-time crop averaging, and ensembles shape what the headline benchmark does and does not prove.

NeurIPS 2012. Alex Krizhevsky, Ilya Sutskever, and Geoffrey E. Hinton. 35 min read, easy difficulty.