Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Gyroscope-Assisted Motion Deblurring Network

Simin Luan, Cong Yang, Zeyd Boukhers, Xue Qin, Dongfeng Cheng, Wei Sui, Zhijun Li

TL;DR

The paper tackles realistic motion blur by addressing the lack of pixel-aligned training triplets and proposing an IMU-driven pipeline to synthesize blur trajectories. It introduces the Gyroscope-Aided Motion Deblurring (GAMD) network, which uses Bezier-encoded blur heatmaps derived from IMU data to guide deblurring within a Feature Pyramid Network framework. The authors present the IMU-blur dataset and BlurTrack for robust evaluation, demonstrating per-pixel trajectory accuracy with about two-pixel error and a PSNR improvement of roughly 33% over the state-of-the-art MIMO method. Overall, the work yields more realistic blur synthesis and superior deblurring performance, enabling scalable data generation and practical deployment in real-world scenarios.

Abstract

Image research has shown substantial attention in deblurring networks in recent years. Yet, their practical usage in real-world deblurring, especially motion blur, remains limited due to the lack of pixel-aligned training triplets (background, blurred image, and blur heat map) and restricted information inherent in blurred images. This paper presents a simple yet efficient framework to synthetic and restore motion blur images using Inertial Measurement Unit (IMU) data. Notably, the framework includes a strategy for training triplet generation, and a Gyroscope-Aided Motion Deblurring (GAMD) network for blurred image restoration. The rationale is that through harnessing IMU data, we can determine the transformation of the camera pose during the image exposure phase, facilitating the deduction of the motion trajectory (aka. blur trajectory) for each point inside the three-dimensional space. Thus, the synthetic triplets using our strategy are inherently close to natural motion blur, strictly pixel-aligned, and mass-producible. Through comprehensive experiments, we demonstrate the advantages of the proposed framework: only two-pixel errors between our synthetic and real-world blur trajectories, a marked improvement (around 33.17%) of the state-of-the-art deblurring method MIMO on Peak Signal-to-Noise Ratio (PSNR).

Gyroscope-Assisted Motion Deblurring Network

TL;DR

The paper tackles realistic motion blur by addressing the lack of pixel-aligned training triplets and proposing an IMU-driven pipeline to synthesize blur trajectories. It introduces the Gyroscope-Aided Motion Deblurring (GAMD) network, which uses Bezier-encoded blur heatmaps derived from IMU data to guide deblurring within a Feature Pyramid Network framework. The authors present the IMU-blur dataset and BlurTrack for robust evaluation, demonstrating per-pixel trajectory accuracy with about two-pixel error and a PSNR improvement of roughly 33% over the state-of-the-art MIMO method. Overall, the work yields more realistic blur synthesis and superior deblurring performance, enabling scalable data generation and practical deployment in real-world scenarios.

Abstract

Image research has shown substantial attention in deblurring networks in recent years. Yet, their practical usage in real-world deblurring, especially motion blur, remains limited due to the lack of pixel-aligned training triplets (background, blurred image, and blur heat map) and restricted information inherent in blurred images. This paper presents a simple yet efficient framework to synthetic and restore motion blur images using Inertial Measurement Unit (IMU) data. Notably, the framework includes a strategy for training triplet generation, and a Gyroscope-Aided Motion Deblurring (GAMD) network for blurred image restoration. The rationale is that through harnessing IMU data, we can determine the transformation of the camera pose during the image exposure phase, facilitating the deduction of the motion trajectory (aka. blur trajectory) for each point inside the three-dimensional space. Thus, the synthetic triplets using our strategy are inherently close to natural motion blur, strictly pixel-aligned, and mass-producible. Through comprehensive experiments, we demonstrate the advantages of the proposed framework: only two-pixel errors between our synthetic and real-world blur trajectories, a marked improvement (around 33.17%) of the state-of-the-art deblurring method MIMO on Peak Signal-to-Noise Ratio (PSNR).
Paper Structure (16 sections, 8 equations, 9 figures, 4 tables)

This paper contains 16 sections, 8 equations, 9 figures, 4 tables.

Table of Contents

  1. Introduction
  2. Related Works
  3. Motion Blur Collection
  4. Deblurring
  5. Method
  6. Blur Trajectories
  7. Deblurring Network
  8. Blur Trajectory in the Network
  9. Network Architecture
  10. Experiments
  11. Datasets
  12. Blur Synthetic Evaluation
  13. Ablation Study
  14. Performance Comparison
  15. Limitations
  16. ...and 1 more sections

Figures (9)

  • Figure 1: Sample results from our proposed framework: The input blurred images (top), and after deblurring (bottom).
  • Figure 2: Synthetic blur images using our strategy. The blurs are variant and close to the real-world, e.g., rotation blur.
  • Figure 3: (a) Obtain blurred image trajectories and corresponding IMU data. (b) Use our strategy to synthesize training triplets. (c) GAMD deblurring network.
  • Figure 4: Coordinate system transformation corresponding to different camera motions of Yaw, Roll, and Pitch.
  • Figure 5: The clear shape image (top), blurred trajectory calculated from the IMU data (middle), and blurred image (bottom) synthesized by the method in this paper.
  • ...and 4 more figures