Writing study: LSRM: High-Fidelity Object-Centric Reconstruction via Scaled Context Windows
← ECCV 2026 oral writing corpus
Zhengqin Li, Cheng Zhang, Jakob Engel, Dong Zhao · 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
- 9 / 4
- 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. Missing split/side-wrapped caption labels restored after rendered-page inspection.
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
Motivates a fine-detail gap between feed-forward reconstruction and dense-view optimization, then targets context-window scale rather than claiming that larger models alone solve it.
- 3 Method: coarse dense stage, sparse stage and parallelism
Describes coarse initialization before sparse high-resolution residuals and workload balancing; experiments link context and routing changes back to texture fidelity.
- 4 Experiments; 5 Conclusions
The source specifies metric rendering resolution, matched input-view counts and hardware, then tests reduced-view inference separately from the main higher-context condition.
Problem, claim and contribution
Motivates a fine-detail gap between feed-forward reconstruction and dense-view optimization, then targets context-window scale rather than claiming that larger models alone solve it.
Abstract and Section 1, pp.1–2 · Main paper
Argument and section order
Describes coarse initialization before sparse high-resolution residuals and workload balancing; experiments link context and routing changes back to texture fidelity.
Sections 3–4, pp.4–15 · Main paper
Figures, captions and visual explanation
The overview draws the coarse and sparse stages separately, including their initialization connection; the main comparison table gives resolution and routing ablations distinct column groups.
Figure 2, p.5; Table 1, p.12 · Main paper
Tables, comparisons and evidence
The source specifies metric rendering resolution, matched input-view counts and hardware, then tests reduced-view inference separately from the main higher-context condition.
Section 4, pp.11–12; Tables 3–4, p.15 · Main paper
Appendix and reproducibility
Background on sparse attention, implementation details and training-stage refinements are referred to supplementary material; those additional pages were not inspected.
Sections 3.1–3.5, pp.5,8,11 · Main paper
Clarity and writing conventions
The method roadmap names initialization, context scaling, fidelity and implementation as separate concerns; limitations avoid claiming demonstrated large-scene generality.
Section 3, pp.4–5; Limitations, p.15 · Main paper
What to reuse—and what not to infer
Transferable lesson: For scaling claims, expose image/volume token counts, the load-balancing strategy and matched-resolution metrics, not just an attractive zoomed crop.
Boundary: Evaluation is object-centric; the framework's potential for large scenes or general generation is a future direction, not a demonstrated result.
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
- 5, 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.