Writing study: SuperDec: 3D Scene Decomposition with Superquadrics Primitives

ICCV overviewCorpus atlasBy contribution typeVisual galleryRules & templateWriting patternsPaper studies

← ICCV 2025 oral writing corpus

Elisabetta Fedele, Boyang Sun, Leonidas Guibas, Marc Pollefeys, Francis Engelmann · 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

≈ 8.04 page-equivalents

Main figures / tables

10 / 3

Attached appendix figures / tables

0 / 0

Attached appendix span

No boundary detected; not assumed zero

Before references / appendix: ≈ 73.09%References / intervening material: ≈ 26.91%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	≈ 1.03 pp.	9.36%
2. Related Work	2	≈ 0.86 pp.	7.82%
3. Method	3	≈ 1.61 pp.	14.64%
4. Experiments	4	≈ 4.03 pp.	36.64%
5. Conclusion	8	≈ 0.24 pp.	2.18%
References	9	≈ 2.96 pp.	26.91%

All detected figure and table locations

Main text

• Figure 1 · p. 1
• Figure 2 · p. 4
• Table 1 · p. 5
• Figure 3 · p. 5
• Table 2 · p. 6
• Table 3 · p. 6
• Figure 4 · p. 6
• Figure 5 · p. 7
• Figure 6 · p. 7
• Figure 7 · p. 7
• Figure 8 · p. 8
• Figure 9 · p. 8
• Figure 10 · p. 8

Attached appendix

No recognized caption labels.

Argument and organization

1. 1 Introduction; 2 Related Work

   Positions superquadrics as compact geometry rather than primarily photorealistic rendering primitives.

2. 3 Method

   Separates learned object-level initialization, geometric refinement and scene-level instance extraction.

3. 4 Experiments; 5 Conclusion

   Progresses from reconstruction and compactness to downstream demonstrations and exploratory representation analyses.

Problem, claim and contribution

The paper makes accuracy and number of primitives a paired objective, motivating shape-expressive primitives with a small parameter overhead.

Sections 1–2, physical pp.1–2 · Main paper

Argument and section order

Scene decomposition is a distinct subsection after the object model, exposing the external Mask3D dependency and per-object normalization.

Sections 3.1–3.2, pp.3–4 · Main paper

Figures, captions and visual explanation

The overview displays tensor dimensions, separate segmentation/parameter heads and an LM refinement stage, making learned initialization distinct from final optimization.

Figure 2, p.4 · Main paper

Tables, comparisons and evidence

Scene reconstruction tables pair Chamfer distances with primitive counts; path-planning results jointly display time, success and memory and retain unfavorable trade-offs.

Tables 2–3, p.6 · Main paper

Appendix and reproducibility

Training details, extra qualitative results, downstream protocols and runtime details are delegated to supplementary material, not inspected.

Sections 4.1–4.3, pp.5–8 · Main paper

Clarity and writing conventions

The analysis uses explicit questions about hierarchy, learned embeddings and speed, while the conclusion calls the representation a first step.

Section 4.3 and conclusion, p.8 · Main paper

What to reuse—and what not to infer

Transferable lesson: Evaluate a compact representation on both its approximation error and its resource footprint, then label downstream demonstrations separately from core quantitative evidence.

Boundary: Object-level self-supervision does not make the full pipeline free of supervised components: scene masks use Mask3D and ScanNet++ object evaluation uses ground-truth masks. The forward-pass time excludes segmentation and iterative LM refinement; selected robot and image-generation demonstrations are not general safety or semantic-understanding validation.

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, 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 · 11 pages