Writing study: ForestFormer3D: A Unified Framework for End-to-End Segmentation of Forest LiDAR 3D Point Clouds

ICCV overviewCorpus atlasBy contribution typeVisual galleryRules & templateWriting patternsPaper studies

← ICCV 2025 oral writing corpus

Binbin Xiang, Maciej Wielgosz, Stefano Puliti, Kamil Král, Martin Krůček, Azim Missarov, Rasmus Astrup · 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: Architecture & representation (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

11 physical pages

Before references / appendix

≈ 8.11 page-equivalents

Main figures / tables

4 / 3

Attached appendix figures / tables

0 / 0

Attached appendix span

No boundary detected; not assumed zero

Before references / appendix: ≈ 73.73%References / intervening material: ≈ 26.27%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	≈ 0.95 pp.	8.64%
2. Related work	2	≈ 1.36 pp.	12.36%
3. Method	3	≈ 2.4 pp.	21.82%
4. Experiments	6	≈ 2.69 pp.	24.45%
5. Conclusion	8	≈ 0.28 pp.	2.55%
Acknowledgements	9	≈ 0.06 pp.	0.55%
References	9	≈ 2.89 pp.	26.27%

All detected figure and table locations

Main text

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

Attached appendix

No recognized caption labels.

Argument and organization

1. 1 Introduction; 2 Related Work

   Motivates forest-specific issues of canopy overlap, tree scale and sensor-dependent point density.

2. 3 Method

   Connects instance-aware query sampling, direct query/tree association and scene-level block merging.

3. 4 Experiments; 5 Conclusion

   Evaluates the expanded forest dataset, two unseen test regions and component contributions to instance and semantic tasks.

Problem, claim and contribution

The initial challenge list is grounded in contrasting forest scenes rather than assuming urban segmentation issues transfer unchanged.

Introduction, physical p.1; Figure 1, p.2 · Main paper

Argument and section order

Training association and inference block merging are distinct subsections, making end-to-end local prediction separate from full-scene assembly.

Sections 3.2–3.3, pp.5–6 · Main paper

Figures, captions and visual explanation

The overview highlights proposed components and explicitly contrasts randomly sampled training cylinders with overlapping inference cylinders.

Figure 2, p.4 · Main paper

Tables, comparisons and evidence

Results jointly expose instance F1/coverage and semantic mIoU; the model is second to TreeLearn on LAUTx instance metrics and does not win every semantic comparison.

Tables 1–2, p.7 · Main paper

Appendix and reproducibility

Dataset statistics, annotation protocols and further factor analysis are routed to supplementary material, not inspected.

Sections 4.1 and 4.3, pp.6–7 · Main paper

Clarity and writing conventions

The text explicitly acknowledges small semantic mIoU reductions while emphasizing the intended gain in individual-tree segmentation.

Section 4.2, p.7; Section 4.3, p.8 · Main paper

What to reuse—and what not to infer

Transferable lesson: For a specialized domain, connect acquisition geometry to model and inference design, and keep competing task outcomes visible in the main table.

Boundary: The TreeLearn FOR-instanceV2 row excludes Yuchen because it fails there, so the table is not entirely a matched-region comparison. Training is local-cylinder based and full scenes require cropping, overlap handling and boundary-mask removal. Transfer to two forests is not unrestricted robustness across sensors, forest types or ecological endpoints.

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

4, 7

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 · 11 pages