%0 Journal Article %1 Krizhevsky_2017 %A Krizhevsky, Alex %A Sutskever, Ilya %A Hinton, Geoffrey E. %D 2017 %I Association for Computing Machinery (ACM) %J Communications of the ACM %K %N 6 %P 84-90 %R 10.1145/3065386 %T ImageNet Classification with Deep Convolutional Neural Networks %U https://doi.org/10.1145%2F3065386 %V 60 %X We trained a large, deep convolutional neural network to classify the 1.2 million high-resolution images in the ImageNet LSVRC-2010 contest into the 1000 different classes. On the test data, we achieved top-1 and top-5 error rates of 37.5 percent and 17.0 percent, respectively, which is considerably better than the previous state-of-the-art. The neural network, which has 60 million parameters and 650,000 neurons, consists of five convolutional layers, some of which are followed by max-pooling layers, and three fully connected layers with a final 1000-way softmax. To make training faster, we used non-saturating neurons and a very efficient GPU implementation of the convolution operation. To reduce overfitting in the fully connected layers we employed a recently developed regularization method called "dropout" that proved to be very effective. We also entered a variant of this model in the ILSVRC-2012 competition and achieved a winning top-5 test error rate of 15.3 percent, compared to 26.2 percent achieved by the second-best entry.

@article{Krizhevsky_2017, abstract = {We trained a large, deep convolutional neural network to classify the 1.2 million high-resolution images in the ImageNet LSVRC-2010 contest into the 1000 different classes. On the test data, we achieved top-1 and top-5 error rates of 37.5 percent and 17.0 percent, respectively, which is considerably better than the previous state-of-the-art. The neural network, which has 60 million parameters and 650,000 neurons, consists of five convolutional layers, some of which are followed by max-pooling layers, and three fully connected layers with a final 1000-way softmax. To make training faster, we used non-saturating neurons and a very efficient GPU implementation of the convolution operation. To reduce overfitting in the fully connected layers we employed a recently developed regularization method called "dropout" that proved to be very effective. We also entered a variant of this model in the ILSVRC-2012 competition and achieved a winning top-5 test error rate of 15.3 percent, compared to 26.2 percent achieved by the second-best entry.}, added-at = {2021-01-17T16:49:04.000+0100}, author = {Krizhevsky, Alex and Sutskever, Ilya and Hinton, Geoffrey E.}, biburl = {https://www.bibsonomy.org/bibtex/2d77365d45cc8b96b898e9c8aae8d0204/peggyschnetter}, doi = {10.1145/3065386}, interhash = {f4eca1b2292406565dd5eec9e305a2f6}, intrahash = {d77365d45cc8b96b898e9c8aae8d0204}, journal = {Communications of the ACM}, keywords = {}, month = may, number = 6, pages = {84-90}, publisher = {Association for Computing Machinery ({ACM})}, timestamp = {2021-01-17T16:49:04.000+0100}, title = {ImageNet Classification with Deep Convolutional Neural Networks}, url = {https://doi.org/10.1145%2F3065386}, volume = 60, year = 2017 }