Writing study: Chorus: Multi-Teacher Pretraining for Holistic 3D Gaussian Scene Encoding

CVPR overviewCorpus atlasBy contribution typeVisual galleryRules & templateWriting patternsPaper studies

← CVPR 2026 oral writing corpus

Yue Li, Qi Ma, Runyi Yang, Mengjiao Ma, Bin Ren, Nikola Popovic, Nicu Sebe, Theo Gevers, Luc Van Gool, Danda Paudel, Martin R. Oswald · Official CVPR 2026 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: Training, objectives & optimization (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

12 physical pages

Before references / appendix

≈ 8.16 page-equivalents

Main figures / tables

8 / 10

Attached appendix figures / tables

0 / 0

Attached appendix span

Separate file recorded below; not in this PDF total

Before references / appendix: ≈ 68%References / intervening material: ≈ 32%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	≈ 0.97 pp.	8.08%
2. Related Work	2	≈ 0.55 pp.	4.58%
3. Method	3	≈ 1.85 pp.	15.42%
4. Experiments	5	≈ 3.61 pp.	30.08%
5. Conclusion	8	≈ 0.21 pp.	1.75%
6. Acknowledgement	9	≈ 0.11 pp.	0.92%
References	9	≈ 3.84 pp.	32%

All detected figure and table locations

Main text

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

Attached appendix

No recognized caption labels.

Separate source files—excluded from canonical-PDF totals

A separate author version may repeat main text and references. Whole-file lengths and counts are not extra appendix material and are never added to the canonical PDF totals. Read the study's source scope for which sections received substantive reading versus layout inspection.

Additional source 1

Separate PDF (or ZIP containing it): whole-file layout of 10 physical pages; 9 detected figures and 3 tables. These may include overlapping main-text material.

Detected heading	PDF page	Estimated span
References	10	≈ 0.95 pp.

• Figure A · p. 2
• Figure B · p. 2
• Figure C · p. 2
• Table A · p. 4
• Figure D · p. 4
• Figure E · p. 5
• Figure F · p. 5
• Table B · p. 7
• Table C · p. 7
• Figure G · p. 7
• Figure H · p. 8
• Figure I · p. 9

Argument and organization

1. Method §§3.1–3.4

   Establish 2D-to-3D supervision, multi-teacher training, rendering adaptation and representation-aware augmentation.

2. Experiments §§4.1–4.3

   Move from native Gaussian tasks to point-cloud transfer, then analyze teachers, costs and design choices.

Problem, claim and contribution

The paper distinguishes rendering a scene from encoding transferable scene features, so its gap is a representation capability rather than an incremental rendering improvement.

Abstract and introduction, PDF pp. 1–2 · Main paper · Additional source 2

Argument and section order

Lifting teacher features is explained before distillation, and online adaptation is a separate subsection; this reveals where offline preprocessing versus online computation occurs.

§§3.1–3.3, PDF pp. 3–4 · Main paper · Additional source 2

Figures, captions and visual explanation

Figure 2 uses an explicit arrow legend for data collection, inference, supervision and online adaptation, and separates pretraining from downstream tasks with a divider.

Figure 2, PDF p. 3 · Main paper · Additional source 2

Tables, comparisons and evidence

Table 1 displays training-source and training-scene columns alongside modality markers; Tables 2–3 separately address instance segmentation and question answering rather than treating a single score as holistic understanding.

Tables 1–3, PDF p. 6 · Main paper · Additional source 2

Appendix and reproducibility

The supplement has a navigable contents list, a rendering-adaptation algorithm, exact teacher feature sizes, dataset split rules, implementation mappings and additional feature projections.

Official ZIP PDF supplement §§A–E, pp. 1–10, Algorithm 1 and Tables A–C · Main paper · Additional source 2

Clarity and writing conventions

Resource claims are broken into preprocessing, uplifting, storage and runtime instead of equating fewer pretraining scenes with universally cheaper training.

Table 7, PDF p. 7; Table 10, PDF p. 8 · Main paper · Additional source 2

What to reuse—and what not to infer

Transferable lesson: A multi-teacher paper benefits from an arrow-type legend and a benchmark table that exposes modality and data volume, followed by explicit teacher and cost ablations.

Boundary: Foreground metrics exclude wall/floor/ceiling and the data-efficiency comparisons involve different pretraining sources; preserve these qualifications when borrowing the headline.

Apply this lesson using the argument and evidence worksheet.

Inspection record and sources

Checked

2026-09-26

Reading status

Writing-focused manuscript scan + selected visuals inspected

Main text

Manuscript-wide writing scan of PDF pp. 1–12: framing, argument sequence, section organization, figure/table captions, evaluation interpretation and limitations; reference sections identified. This is a writing-focused scan, not an exhaustive technical proof or reproduction audit. Main PDF pages 3 and 6 were rendered and visually inspected.

Appendix / supplement

Writing-focused scan of the separate supplementary PDF, pp. 1–10, including its organization, captions, protocol/evidence detail and limitation material where present; proofs and code listings were inspected for exposition, not independently verified or executed. Official ZIP PDF supplement §§A–E, pp. 1–10, Algorithm 1 and Tables A–C. Linked video/audio demonstrations and non-PDF archive contents were not assessed.

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 CVPR 2026 CVF open-access accepted manuscript; retrieved 2026-09-26 · 12 pages
• Official CVPR 2026 supplementary ZIP PDF member: CVPR_2026_Chorus_supplement.pdf; retrieved 2026-09-26 · 10 pages