2012, in 5 pieces
5 records · 0 notes
December 1
Alex Krizhevsky, Ilya Sutskever, Geoffrey Hinton
ImageNet Classification with Deep Convolutional Neural Networks (AlexNet)
Established in a paper the combination of GPUs, large-scale data and deep CNNs. The conventional starting point of the current deep learning era
October 1
SuperVision (Krizhevsky, Sutskever, Hinton, University of Toronto)
SuperVision wins ILSVRC 2012 with a 15.3% top-5 error rate
The turning point at which, with a wide margin over second place, deep CNNs became the mainstream methodology of computer vision
July 1
Geoffrey Hinton et al. (University of Toronto)
Improving neural networks by preventing co-adaptation of feature detectors
Curbed overfitting with simple regularization and later became a standard technique
June 1
Yoshua Bengio, Aaron Courville, Pascal Vincent
Representation Learning: A Review and New Perspectives
Framed representation learning as a single problem, becoming the field's standard reference from then on
Year only 1
Geoffrey Hinton, Li Deng, Dong Yu et al. (University of Toronto, Microsoft, Google, IBM)
Deep Neural Networks for Acoustic Modeling in Speech Recognition
Separate industry groups confirmed the same shift (HMM→DNN), and deep learning became the industry standard for speech recognition