SORT: Second-Order Response Transform for Visual Recognition
This addresses the need for more flexible and nonlinear network functions in visual recognition, though it appears incremental as it builds on existing two-branch architectures.
The paper tackles the problem of enhancing deep neural networks by introducing second-order operations, proposing SORT to add element-wise product transforms to two-branch modules, resulting in consistent accuracy gains on datasets like CIFAR10, CIFAR100, SVHN, and ILSVRC2012 with less than 5% extra computation overhead.
In this paper, we reveal the importance and benefits of introducing second-order operations into deep neural networks. We propose a novel approach named Second-Order Response Transform (SORT), which appends element-wise product transform to the linear sum of a two-branch network module. A direct advantage of SORT is to facilitate cross-branch response propagation, so that each branch can update its weights based on the current status of the other branch. Moreover, SORT augments the family of transform operations and increases the nonlinearity of the network, making it possible to learn flexible functions to fit the complicated distribution of feature space. SORT can be applied to a wide range of network architectures, including a branched variant of a chain-styled network and a residual network, with very light-weighted modifications. We observe consistent accuracy gain on both small (CIFAR10, CIFAR100 and SVHN) and big (ILSVRC2012) datasets. In addition, SORT is very efficient, as the extra computation overhead is less than 5%.