“DanNet: Flexible, High Performance Convolutional Neural Networks for Image Classification”, Dan Claudiu Ciresan, Ueli Meier, Jonathan Masci, Luca Maria Gambardella, Jürgen Schmidhuber2011-02-01 (CNN, AI hardware)⁠:

[followup: Cireşan et al 2012a/b/c] We present a fast, fully parameterizable GPU implementation of Convolutional Neural Network variants. Our feature extractors are neither carefully designed nor pre-wired, but rather learned in a supervised way.

Our deep hierarchical architectures achieve the best published results on benchmarks for object classification (NORB, CIFAR-10) and handwritten digit recognition (MNIST), with error rates of 2.53%, 19.51%, 0.35%, respectively. Deep nets trained by simple back-propagation perform better than more shallow ones.

Learning is surprisingly rapid. NORB is completely trained within 5 epochs. Test error rates on MNIST drop to 2.42%, 0.97% and 0.48% after 1, 3 and 17 epochs, respectively.

[Prior to AlexNet; retrospective]