Writing study: Image as an IMU: Estimating Camera Motion from a Single Motion-Blurred Image

ICCV overviewCorpus atlasBy contribution typeVisual galleryRules & templateWriting patternsPaper studies

← ICCV 2025 oral writing corpus

Jerred Chen, Ronald Clark · Official ICCV 2025 oral record

Observed writing choices, not acceptance rules. This study examines the inspected manuscript version. It does not establish why the paper was accepted or that its presentation choices should be copied unchanged.

Figure, table and section ledger

Editorial contribution type: Domain-grounded methods & physical systems (primary).

Layout extraction, not a required paper template. Counts are distinct caption labels; a multi-panel figure counts once. Section spans are estimated from heading positions, including floats and whitespace. Method and limitations.

Canonical PDF

10 physical pages

Before references / appendix

≈ 8.05 page-equivalents

Main figures / tables

7 / 3

Attached appendix figures / tables

0 / 0

Attached appendix span

No boundary detected; not assumed zero

Before references / appendix: ≈ 80.5%References / intervening material: ≈ 19.5%Estimated layout envelopes, not text/artwork area. Unresolved appendix boundaries remain within the undivided post-body material.

Extraction notes: No attached appendix boundary detected. Zero means no recognized caption, not independently verified absence.

Section-by-section lengths and source pages

The title/abstract/front matter precede the first numbered heading. A span ends at the next detected top-level heading, not at the last sentence; end-page whitespace is included. References and statements are listed separately. These are not measurements of text-only or artwork area.

Top-level heading	Starts on PDF page	Estimated span	Share of entire PDF
1. Introduction	1	≈ 0.9 pp.	9%
2. Related Works	2	≈ 0.68 pp.	6.8%
3. Preliminaries	2	≈ 0.85 pp.	8.5%
4. Method	3	≈ 1.54 pp.	15.4%
5. Dataset Curation	5	≈ 1.1 pp.	11%
6. Experiments	6	≈ 2.46 pp.	24.6%
7. Conclusion	8	≈ 0.18 pp.	1.8%
References	9	≈ 1.95 pp.	19.5%

All detected figure and table locations

Main text

• Figure 1 · p. 1
• Figure 2 · p. 3
• Figure 3 · p. 5
• Table 1 · p. 6
• Table 2 · p. 6
• Figure 4 · p. 7
• Figure 5 · p. 8
• Figure 6 · p. 8
• Figure 7 · p. 8
• Table 3 · p. 8

Attached appendix

No recognized caption labels.

Argument and organization

1. 1 Introduction; 2 Related Works; 3 Preliminaries

   Recasts exposure blur as virtual correspondences, then introduces image formation before the estimator.

2. 4 Method; 5 Dataset Curation

   Separates flow/depth prediction, differentiable velocity fitting and direction disambiguation; synthetic and real data have distinct provenance.

3. 6 Experiments; 7 Conclusion

   Reports per-axis errors, trajectory diagnostics, robotic capture, IMU integration comparison and runtime.

Problem, claim and contribution

Turns an unwanted artifact into a motion cue while specifying small-motion and rigid-scene assumptions for velocity recovery.

Abstract and Sections 3–4, physical pp.1,3–4 · Main paper

Argument and section order

The image-formation explanation comes before the learned front end and least-squares stage; directional ambiguity gets its own subsection.

Sections 3–4.3, pp.3–5 · Main paper

Figures, captions and visual explanation

The overview separates learned flow/depth outputs from the explicit linear-system solve and labels tensor shapes and angular/translational outputs.

Figure 2, p.3 · Main paper

Tables, comparisons and evidence

Per-axis tables separate rotational rad/s from translational m/s, distinguish multi-image and single-image methods, retain failed reconstructions as crosses and include a zero-velocity baseline.

Tables 1–2, p.6 · Main paper

Appendix and reproducibility

The small-motion derivation and additional severe-blur examples are referred to supplementary material; those pointers alone were inspected.

Section 4.2, p.4; Section 6, p.8 · Main paper

Clarity and writing conventions

The data section names ARKit and PromptDA-derived quantities rather than implying motion-capture truth, and explicitly limits training to indoor scenes.

Sections 5.1–5.2, pp.5–6 · Main paper

What to reuse—and what not to infer

Transferable lesson: For inverse problems, place ambiguity resolution beside the main method and distinguish measured supervision from model-derived references.

Boundary: The single-frame estimator has a 180-degree direction ambiguity; its proposed sign-resolution heuristic uses neighboring frames. Angular evaluation uses gyroscope measurements and translation uses ARKit-derived velocities, so an IMU comparison is not independent proof of unrestricted sensor replacement.

Apply this lesson using the argument and evidence worksheet.

Inspection record and sources

Checked

2026-10-07

Reading status

Writing-focused manuscript scan + selected visuals inspected

Main text

Writing-oriented scan across the main manuscript, physical PDF pp.1–8, covering framing, section progression, method explanations, captions, comparison protocols and concluding scope; reference boundary checked. Selected full-page rasters were inspected. Not exhaustive line-by-line reading, proof verification or experimental reproduction.

Appendix / supplement

Main-paper appendix/supplement references were checked. A separate supplementary source is linked in the inventory but was not retrieved or inspected; its contents are not included in the reading or visual-page counts.

Visually inspected PDF pages

3, 6

This is a writing-focused scan of the manuscript and the stated supplementary scope: framing, organization, visual explanation, evidence presentation and wording. It is not a line-by-line proof audit, reproduction of results or exhaustive scientific review.

• Official CVF open-access ICCV 2025 accepted-paper PDF, retrieved 7 October 2026; CVF states identical to accepted version except watermark; not assumed identical to initial submission · 10 pages