The context to this post are recent papers providing evidence that prediction performance of deep networks trained for image classification is in majority tue to texture biases, i.e. networks take decisions based on textures / small patches : Geirhos et al., ImageNet-trained CNNs are biased towards texture; increasing shape bias…

Multi-layer Perceptrons (MLPs) are standard neural networks with fully connected layers, where each input unit is connected with each output unit. We rarely use these kind of layers when working with images, at least not in the early stages of processing. Why is this so? …

Caveat: I wrote this story initially on Dec. 21st, 2016, but moved it to medium in Oct. 2018. Why is the result in the ICLR 2017 best paper “Understanding deep learning requires rethinking generalization” so interesting? (https://arxiv.org/abs/1611.03530) Let’s think about model fitting and generalization using an example from Christopher Bishop’s…