Motion Deblurring with Real Events

TL;DR

This paper introduces a self-supervised, end-to-end learning framework for event-based motion deblurring that leverages real-world events and a piece-wise linear motion model to improve deblurring accuracy in real-world scenarios.

Contribution

It presents a novel self-supervised approach using real-world events and a new motion model to enhance event-based motion deblurring performance.

Findings

Achieves superior deblurring on synthetic and real datasets.

Bridges the gap between simulated and real-world motion blurs.

Demonstrates robustness in real-world motion deblurring tasks.

Abstract

In this paper, we propose an end-to-end learning framework for event-based motion deblurring in a self-supervised manner, where real-world events are exploited to alleviate the performance degradation caused by data inconsistency. To achieve this end, optical flows are predicted from events, with which the blurry consistency and photometric consistency are exploited to enable self-supervision on the deblurring network with real-world data. Furthermore, a piece-wise linear motion model is proposed to take into account motion non-linearities and thus leads to an accurate model for the physical formation of motion blurs in the real-world scenario. Extensive evaluation on both synthetic and real motion blur datasets demonstrates that the proposed algorithm bridges the gap between simulated and real-world motion blurs and shows remarkable performance for event-based motion deblurring in real-world scenarios.