Improved Handling of Motion Blur in Online Object Detection

TL;DR

This paper introduces novel methods to improve online object detection in motion-blurred images, especially addressing egomotion-induced blur, resulting in significantly enhanced detection performance in real-world scenarios.

Contribution

It proposes specific remedies including custom label generation and blur-type conditioning, which substantially improve detection accuracy over existing approaches.

Findings

Custom label generation markedly improves detection.

Conditioning on blur type boosts detection in motion-blurred images.

Combined remedies outperform previous methods on real-world datasets.

Abstract

We wish to detect specific categories of objects, for online vision systems that will run in the real world. Object detection is already very challenging. It is even harder when the images are blurred, from the camera being in a car or a hand-held phone. Most existing efforts either focused on sharp images, with easy to label ground truth, or they have treated motion blur as one of many generic corruptions. Instead, we focus especially on the details of egomotion induced blur. We explore five classes of remedies, where each targets different potential causes for the performance gap between sharp and blurred images. For example, first deblurring an image changes its human interpretability, but at present, only partly improves object detection. The other four classes of remedies address multi-scale texture, out-of-distribution testing, label generation, and conditioning by blur-type. Surprisingly, we discover that custom label generation aimed at resolving spatial ambiguity, ahead of all others, markedly improves object detection. Also, in contrast to findings from classification, we see a noteworthy boost by conditioning our model on bespoke categories of motion blur. We validate and cross-breed the different remedies experimentally on blurred COCO images and real-world blur datasets, producing an easy and practical favorite model with superior detection rates.