This paper studies if crowding, a visual effect suffered by human visual systems, happens to deep neural network as well. The paper systematically analyzes the performance difference when (1) clutter/flankers is present; (2) the similarity and proximity to the target; (3) when different architectures of the network is used. Pros: There are very few papers to study if various visual perceptual phenomenon exists in deep neural nets, or in vision algorithms in general. This paper studies the effect of crowding in DNN/DCNN image classification problem, and presents some interesting results which seems to suggest similar effect exists in DNN because of pooling layers merges nearby responses. And this is related to the theories of crowding in humans, which is also interesting.

Crowding is a visual effect suffered by humans, in which an object that can be recognized in isolation can no longer be recognized when other objects, called flankers, are placed close to it. In this work, we study the effect of crowding in artificial Deep Neural Networks (DNNs) for object recognition. We analyze both deep convolutional neural networks (DCNNs) as well as an extension of DCNNs that are multi-scale and that change the receptive field size of the convolution filters with their position in the image. The latter networks, that we call eccentricity-dependent, have been proposed for modeling the feedforward path of the primate visual cortex. Our results reveal that the eccentricity-dependent model, trained on target objects in isolation, can recognize such targets in the presence of flankers, if the targets are near the center of the image, whereas DCNNs cannot.