A Simple Domain Shifting Networkfor Generating Low Quality Images

TL;DR

This paper introduces a simple domain shifting network that degrades high-quality images to mimic low-quality robot camera images, improving classification accuracy on low-quality data without complex adaptation methods.

Contribution

The paper presents a novel quality degrading network that enhances training for low-quality robot images, outperforming traditional zero-shot domain adaptation techniques.

Findings

Classification accuracy improves with the degrading network training.

The method outperforms zero-shot domain adaptation on real robot data.

Easier to implement than existing domain adaptation approaches.

Abstract

Deep Learning systems have proven to be extremely successful for image recognition tasks for which significant amounts of training data is available, e.g., on the famous ImageNet dataset. We demonstrate that for robotics applications with cheap camera equipment, the low image quality, however,influences the classification accuracy, and freely available databases cannot be exploited in a straight forward way to train classifiers to be used on a robot. As a solution we propose to train a network on degrading the quality images in order to mimic specific low quality imaging systems. Numerical experiments demonstrate that classification networks trained by using images produced by our quality degrading network along with the high quality images outperform classification networks trained only on high quality data when used on a real robot system, while being significantly easier to use than competing zero-shot domain adaptation techniques.