Writing study: Geometric Context Transformer for Streaming 3D Reconstruction
← ECCV 2026 oral writing corpus
Lin-Zhuo Chen, Jian Gao, Shangzhan Zhang, Yihang Chen, Nan Xue, Jianyuan Wang, Christian Rupprecht, Xun Cao, Xing Zhu, Yujun Shen, Yao Yao, YINGHAO XU · Official ECCV 2026 oral record
Figure, table and section ledger
Editorial contribution type: Efficiency, inference & 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
- 21 physical pages
- Before references / appendix
- ≈ 15.15 page-equivalents
- Main figures / tables
- 5 / 7
- Attached appendix figures / tables
- 0 / 0
- Attached appendix span
- No boundary detected; not assumed zero
Extraction notes: No attached appendix boundary detected. Zero means no recognized caption, not independently verified absence.
Source-bound correction: All top-level main headings and the reference boundary checked against source PDF text; source-bound positions use physical single-column y/page-height, not a two-column reading-order surrogate. Running page-number/author titles, affiliations, numbered contribution bullets, footnotes and bibliography continuations are excluded. This linked manuscript contains no appended supplementary section; separate supplement links were not scanned.
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.
All detected figure and table locations
Main text
Attached appendix
No recognized caption labels.
Argument and organization
- 1 Introduction; 2 Related Work
Frames streaming reconstruction as a tension among geometric accuracy, temporal consistency and state cost, rather than only the ability to process another frame.
- 3 Method: geometric context, framework and inference
Defines causal input access, decomposes context into anchor/local/trajectory parts, then explains cached inference before evaluating trajectories, reconstruction and costs.
- 4 Experiments and analysis; 5 Conclusion
Tables label offline, online and optimization-based methods, while trajectory overlays share ground-truth/prediction colors; a window/anchor sweep reports accuracy and FPS together.
Problem, claim and contribution
Frames streaming reconstruction as a tension among geometric accuracy, temporal consistency and state cost, rather than only the ability to process another frame.
Abstract and Section 1, pp.1–3 · Main paper
Argument and section order
Defines causal input access, decomposes context into anchor/local/trajectory parts, then explains cached inference before evaluating trajectories, reconstruction and costs.
Sections 3–4, pp.5–15 · Main paper
Figures, captions and visual explanation
A matched four-mask panel compares full, causal, sliding-window and geometric-context attention, with the same token blocks and explicit context markers across alternatives.
Figure 2, p.7 · Main paper
Tables, comparisons and evidence
Tables label offline, online and optimization-based methods, while trajectory overlays share ground-truth/prediction colors; a window/anchor sweep reports accuracy and FPS together.
Tables 1,3 and 7, pp.11–12,15; Figure 4, p.12 · Main paper
Appendix and reproducibility
The first qualitative figure points to additional supplementary demonstrations, but no separate supplement was retrieved; the main manuscript retains attention design and cost analysis.
Figure 1, p.2; Section 3.2, pp.6–7 · Main paper
Clarity and writing conventions
The overview states that future observations are unavailable, making the streaming contract inspectable before comparing results with offline systems.
Section 3.1, p.5 · Main paper
What to reuse—and what not to infer
Transferable lesson: Streaming papers need a future-access contract, a state/mask diagram and joint accuracy-memory-runtime reporting as sequence length grows.
Boundary: Trajectory-memory tokens still grow with sequence length; 'nearly constant' practical costs should not be rewritten as a strict asymptotic constant-memory theorem.
Apply this lesson using the argument and evidence worksheet.
Inspection record and sources
- Checked
- 2026-10-02
- Reading status
- Writing-focused manuscript scan + selected visuals inspected
- Main text
- Writing-oriented scan across the manuscript main argument, physical PDF pp.1–15: framing and section roadmaps, method/argument explanations, figure and table captions, comparison protocols, conclusions and limitations. Selected full-page rasters were inspected. This is not exhaustive line-by-line technical reading, proof verification or experimental reproduction.
- Appendix / supplement
- Main-manuscript pointers to appendix/supplement and the reference boundary were checked. Separate supplementary PDFs were not retrieved or inspected; their contents are not included in inspected-page counts.
- Visually inspected PDF pages
- 4, 7, 12
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.