Convolutional Neural Networks Regularized by Correlated Noise
This work addresses robustness in image classification for computer vision applications, but it is incremental as it builds on existing regularization techniques like dropout.
The paper tackled the problem of improving convolutional neural network robustness to occluded images by incorporating correlated noise models inspired by the visual cortex, achieving the best performance in 10 out of 12 cases studied.
Neurons in the visual cortex are correlated in their variability. The presence of correlation impacts cortical processing because noise cannot be averaged out over many neurons. In an effort to understand the functional purpose of correlated variability, we implement and evaluate correlated noise models in deep convolutional neural networks. Inspired by the cortex, correlation is defined as a function of the distance between neurons and their selectivity. We show how to sample from high-dimensional correlated distributions while keeping the procedure differentiable, so that back-propagation can proceed as usual. The impact of correlated variability is evaluated on the classification of occluded and non-occluded images with and without the presence of other regularization techniques, such as dropout. More work is needed to understand the effects of correlations in various conditions, however in 10/12 of the cases we studied, the best performance on occluded images was obtained from a model with correlated noise.