F. Li, N. Neverova, C. Wolf, and G. Taylor, “Modout: Learning to Fuse Modalities via Stochastic Regularization”, J. Comp. Vis. Imag. Sys., vol. 2, no. 1, Oct. 2016.