You want to train your deep neural network well ? drop entire layers when training: Think of stochastic depth as a regularizer !
Deep Networks with Stochastic Depth by Gao Huang, Yu Sun, Zhuang Liu, Daniel Sedra, Kilian Weinberger
An implementation can be found here: https://github.com/yueatsprograms/Stochastic_Depth
Very deep convolutional networks with hundreds or more layers have lead to significant reductions in error on competitive benchmarks like the ImageNet or COCO tasks. Although the unmatched expressiveness of the many deep layers can be highly desirable at test time, training very deep networks comes with its own set of challenges. The gradients can vanish, the forward flow often diminishes and the training time can be painfully slow even on modern computers. In this paper we propose stochastic depth, a training procedure that enables the seemingly contradictory setup to train short networks and obtain deep networks. We start with very deep networks but during training, for each mini-batch, randomly drop a subset of layers and bypass them with the identity function. The resulting networks are short (in expectation) during training and deep during testing. Training Residual Networks with stochastic depth is compellingly simple to implement, yet effective. We show that this approach successfully addresses the training difficulties of deep networks and complements the recent success of Residual and Highway Networks. It reduces training time substantially and improves the test errors on almost all data sets significantly (CIFAR-10, CIFAR-100, SVHN). Intriguingly, we show that with stochastic depth we can increase the depth of residual networks even beyond 1200 layers and still yield meaningful improvements in test error (4.91%) on CIFAR-10.
An implementation can be found here: https://github.com/yueatsprograms/Stochastic_Depth
Liked this entry ? subscribe to Nuit Blanche's feed, there's more where that came from. You can also subscribe to Nuit Blanche by Email, explore the Big Picture in Compressive Sensing or the Matrix Factorization Jungle and join the conversations on compressive sensing, advanced matrix factorization and calibration issues on Linkedin.
No comments:
Post a Comment