Deep Neural Networks with Box Convolutions

NeurIPS 2018 Egor Burkov • Victor Lempitsky

Box filters computed using integral images have been part of the computer vision toolset for a long time. Here, we show that a convolutional layer that computes box filter responses in a sliding manner can be used within deep architectures, whereas the dimensions and the offsets of the sliding boxes in such a layer can be learned as part of an end-to-end loss minimization. Crucially, the training process can make the size of the boxes in such a layer arbitrarily large without incurring extra computational cost and without the need to increase the number of learnable parameters.

PDF Abstract

Evaluation


No evaluation results yet. Help compare this paper to other papers by submitting the tasks and evaluation metrics from the paper.