Self-Supervised Linear Motion Deblurring
Journal article, 2020

Motion blurry images challenge many computer vision algorithms, e.g., feature detection, motion estimation, or object recognition. Deep convolutional neural networks are stateof-the-art for image deblurring. However, obtaining training data with corresponding sharp and blurry image pairs can be difficult. In this letter, we present a differentiable reblur model for selfsupervised motion deblurring, which enables the network to learn from real-world blurry image sequences without relying on sharp images for supervision. Our key insight is that motion cues obtained from consecutive images yield sufficient information to inform the deblurring task. We therefore formulate deblurring as an inverse rendering problem, taking into account the physical image formation process: we first predict two deblurred images from which we estimate the corresponding optical flow. Using these predictions, we re-render the blurred images and minimize the difference with respect to the original blurry inputs. We use both synthetic and real dataset for experimental evaluations. Our experiments demonstrate that self-supervised single image deblurring is really feasible and leads to visually compelling results. Both the code and datasets are available at https://github.com/ethliup/SelfDeblur https://github.com/ethliup/SelfDeblur.

Author

Peidong Liu

ETH Zurich

Joel Janai

University of Tübingen

Marc Pollefeys

Swiss Federal Institute of Technology in Zürich (ETH)

Torsten Sattler

Chalmers, Electrical Engineering, Signalbehandling och medicinsk teknik, Imaging and Image Analysis

Andreas Geiger

University of Tübingen

IEEE Robotics and Automation Letters

2377-3766 (ISSN)

Vol. 5 2 2475-2482

Subject Categories

Robotics

DOI

10.1109/LRA.2020.2972873

More information

Latest update

7/20/2020