Fast Image Scanning with Deep Max-Pooling Convolutional Neural Networks

TL;DR

This paper presents a method to significantly accelerate image scanning with deep max-pooling CNNs by applying dynamic programming, making the process more computationally efficient without sacrificing accuracy.

Contribution

The authors introduce a dynamic programming approach that speeds up sliding window image scanning using deep max-pooling CNNs, addressing computational bottlenecks.

Findings

Speedup of image scanning by orders of magnitude

Effective handling of max-pooling layers in dynamic programming

Potential for real-time image analysis applications

Abstract

Deep Neural Networks now excel at image classification, detection and segmentation. When used to scan images by means of a sliding window, however, their high computational complexity can bring even the most powerful hardware to its knees. We show how dynamic programming can speedup the process by orders of magnitude, even when max-pooling layers are present.