# GeoBlur: Epipolar Geometry Estimation from a Single Motion-Blurred Image

arXiv: https://arxiv.org/abs/2605.01552 · HTML: https://mlsanity.com/paper/2605.01552/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 23 of its 52 references that are in the map.
- First posted on arXiv 2 May 2026.

## Builds on (most cited first, 23 of 23)

- [Simple and Scalable Predictive Uncertainty Estimation using Deep Ensembles](https://mlsanity.com/paper/1612.01474/) (2016)
- [SwinIR: Image Restoration Using Swin Transformer](https://mlsanity.com/paper/2108.10257/) (2021)
- [DUSt3R: Geometric 3D Vision Made Easy](https://mlsanity.com/paper/2312.14132/) (2023)
- [FlowNet: Learning Optical Flow with Convolutional Networks](https://mlsanity.com/paper/1504.06852/) (2015)
- [PWC-Net: CNNs for Optical Flow Using Pyramid, Warping, and Cost Volume](https://mlsanity.com/paper/1709.02371/) (2017)
- [A Large Dataset to Train Convolutional Networks for Disparity, Optical Flow, and Scene Flow Estimation](https://mlsanity.com/paper/1512.02134/) (2015)
- [Deep Multi-scale Convolutional Neural Network for Dynamic Scene Deblurring](https://mlsanity.com/paper/1612.02177/) (2016)
- [Super SloMo: High Quality Estimation of Multiple Intermediate Frames for Video Interpolation](https://mlsanity.com/paper/1712.00080/) (2017)
- [RoMa: Robust Dense Feature Matching](https://mlsanity.com/paper/2305.15404/) (2023)
- [Global Structure-from-Motion Revisited](https://mlsanity.com/paper/2407.20219/) (2024)
- [Learning Less is More - 6D Camera Localization via 3D Surface Regression](https://mlsanity.com/paper/1711.10228/) (2017)
- [Uncertainty Estimates and Multi-Hypotheses Networks for Optical Flow](https://mlsanity.com/paper/1802.07095/) (2018)
- [Learning Feature Descriptors using Camera Pose Supervision](https://mlsanity.com/paper/2004.13324/) (2020)
- [From Motion Blur to Motion Flow: a Deep Learning Solution for Removing Heterogeneous Motion Blur](https://mlsanity.com/paper/1612.02583/) (2016)
- [FusionSeg: Learning to combine motion and appearance for fully automatic segmention of generic objects in videos](https://mlsanity.com/paper/1701.05384/) (2017)
- [Bringing Alive Blurred Moments](https://mlsanity.com/paper/1804.02913/) (2018)
- [Learning to Synthesize Motion Blur](https://mlsanity.com/paper/1811.11745/) (2018)
- [Blur Interpolation Transformer for Real-World Motion from Blur](https://mlsanity.com/paper/2211.11423/) (2022)
- [Uncertainty Quantification Metrics for Deep Regression](https://mlsanity.com/paper/2405.04278/) (2024)
- [Deep Fundamental Matrix Estimation without Correspondences](https://mlsanity.com/paper/1810.01575/) (2018)
- [Optical Flow Estimation from a Single Motion-blurred Image](https://mlsanity.com/paper/2103.02996/) (2021)
- [Image as an IMU: Estimating Camera Motion from a Single Motion-Blurred Image](https://mlsanity.com/paper/2503.17358/) (2025)
- [MONET: Multiview Semi-supervised Keypoint Detection via Epipolar Divergence](https://mlsanity.com/paper/1806.00104/) (2018)

## Abstract (arXiv)

Relative camera pose geometry, formulated via fundamental matrix estimation, is a challenging problem in many robotics and VR/AR applications. These applications occasionally contain fast monocular camera motion, which severely blurs the image and prevents the use of traditional multi-view geometry methods for camera pose estimation. To handle these cases, we propose GeoBlur, a framework for estimating the fundamental matrix and recovering relative camera pose directly from a single motion-blurred image, using the motion cues from blur artifacts. GeoBlur first predicts the visual correspondences between two time instances within the camera exposure window; then, it infers the fundamental matrix by solving the single-frame epipolar geometry problem under time-direction ambiguity. The resulting fundamental matrix is unique up to transposition, reflecting the inherent ambiguity in the direction of time. GeoBlur improves performance on synthetic and hybrid benchmarks while remaining competitive with prior work on real motion-blur data. We further demonstrate the use of GeoBlur on the downstream task on single frame motion segmentation.