Research map / 3D vision / Depth and stereo
Image as an IMU: Estimating Camera Motion from a Single Motion-Blurred Image
Jerred Chen, Ronald Clark
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| estimate velocity from a single blurred image instead of multi-image MASt3R | single-image blur method · Real-world test sequences (avg of 4) | rotational velocity RMSE w_x (rad/s): 4.04 -> 1.22 | Improved |
| estimate velocity from a single blurred image instead of multi-image MASt3R | single-image blur method · Real-world test sequences (avg of 4) | translational velocity RMSE v_x (m/s): 1.60 -> 1.11 | Improved |
| compare with second-best method on rotational velocity | single-image blur method · Real-world test sequences | rotational velocity RMSE: 31% reduction | Improved |
| compare with second-best method on translational velocity | single-image blur method · Real-world test sequences | translational velocity RMSE: 24% reduction | Improved |
| compare with zero-velocity prior on translation z | single-image blur method · Real-world test sequences (avg of 4) | translational velocity RMSE v_z (m/s): 1.24 -> 0.92 | Improved |
| COLMAP on blurred sequences | COLMAP · Real-world test sequences | pose reconstruction | Hurt |
| single-image network instead of DROID-SLAM for runtime | single-image blur method · Real-world test sequences | average FPS: 4.34 -> 30.17 | Improved |
| use direct velocity prediction instead of IMU preintegration | single-image blur method · long video | velocity drift: drift after 20 seconds | Improved |
| MASt3R beats ours on dining sequence translation | MASt3R · dining sequence | translational velocity RMSE v_x (m/s): 0.82 vs 0.95 | Mixed |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- DROID-SLAM — DROID-SLAM: Deep Visual SLAM for Monocular, Stereo, and RGB-D Cameras (NeurIPS 2021) · better on 7 of 7 shared results
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Optical Flow Estimation from a Single Motion-blurred Image (AAAI 2021 · similarity 0.89)
- From Motion Blur to Motion Flow: A Deep Learning Solution for Removing Heterogeneous Motion Blur (CVPR 2017 · similarity 0.89)
- Parametric Object Motion From Blur (CVPR 2016 · similarity 0.88)
- Motion-From-Blur: 3D Shape and Motion Estimation of Motion-Blurred Objects in Videos (CVPR 2022 · similarity 0.88)
- MotionPro: A Precise Motion Controller for Image-to-Video Generation (CVPR 2025 · similarity 0.86)
- Zero-Shot Monocular Scene Flow Estimation in the Wild (CVPR 2025 · similarity 0.86)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
