Wired Magazine writes about the history of convolutional nets. Fro some reason, Wired used this picture of me, taken at the 2013 pre-NIPS CIFAR workshop on Neural Computation and Adaptive Perception. I look a bit stern on this picture (Ed Lazowska would say I look like my wallet just got stolen). Actually, Im listening to a talk and probably trying to figure out what the speaker is saying. Regardless, this piece just published in Wired Magazine and authored by Daniela Hernandez talks about the history of convolutional nets, and about some of my career. There are quotes from my friends and colleagues Leon Bottou, Geoff Hinton, Larry Jackel, Yoshua Bengio, and Andrew Ng. Its difficult to write such an article about a piece of science or technology and do justice to the contributions of everyone involved. Ideas do not appear in a vacuum, and I have been fortunate to work with wonderfully talented people in the heydays of convolutional nets, including Leon, Geoff, Larry and Yoshua, and many others at Bell Labs such as John Denker, Rich Howard, Isabelle Guyon, Vladimir Vapnik, Sara Solla, Chris Burges, Patrick Haffner, Ofer Matan, Craig Nohl.... AI research is going through a renaissance, and the excitement I feel at Facebook AI Research reminds me of these early days at Bell Labs. I am fortunate again to be surrounded by stupendously talented people. The article could be interpreted as saying that Leon and I came up with stochastic gradient optimization. We didnt. But we have used it since day one, and we played a role in popularizing it in recent years (though that was mostly Leons doing). #convnet #deeplearning #FacebookAIresearch wired/2014/08/deep-learning-yann-lecun/
Posted on: Thu, 14 Aug 2014 22:23:52 +0000
Recently Viewed Topics
© 2015