Writing study: GT-Loc: Unifying When and Where in Images Through a Joint Embedding Space

ICCV overviewCorpus atlasBy contribution typeVisual galleryRules & templateWriting patternsPaper studies

← ICCV 2025 oral writing corpus

David G. Shatwell, Ishan Rajendrakumar Dave, Sirnam Swetha, Mubarak Shah · 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: 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

11 physical pages

Before references / appendix

≈ 8.14 page-equivalents

Main figures / tables

5 / 6

Attached appendix figures / tables

0 / 0

Attached appendix span

No boundary detected; not assumed zero

Before references / appendix: ≈ 74%References / intervening material: ≈ 26%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	≈ 1.29 pp.	11.73%
2. Related Work	2	≈ 1.1 pp.	10%
3. Method	3	≈ 2.47 pp.	22.45%
4. Experiments	6	≈ 2.61 pp.	23.73%
5. Conclusion	8	≈ 0.22 pp.	2%
6. Acknowledgements	9	≈ 0.1 pp.	0.91%
References	9	≈ 2.86 pp.	26%

All detected figure and table locations

Main text

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

Attached appendix

No recognized caption labels.

Argument and organization

1. 1 Introduction; 2 Related Work

   Establishes that time-of-day and season depend on location, then contrasts metadata requirements and classification versus retrieval.

2. 3 Method

   Defines three encoders, location alignment, cyclic time representation and temporal metric learning before specifying inference.

3. 4 Experiments; 5 Conclusion

   Separates time prediction, geolocation, loss ablations, data/noise robustness and compositional retrieval.

Problem, claim and contribution

Makes the input contract central: jointly estimate month/hour and location from an image, without supplying GPS at inference.

Abstract and Sections 1–2, physical pp.1–3 · Main paper

Argument and section order

Places the cyclic month-hour representation before the loss; inference then reuses the shared embedding notation.

Sections 3.2–3.4, p.5 · Main paper

Figures, captions and visual explanation

The overview uses three modality colors and frozen/trainable icons; the inset explains why a wrapped temporal distance differs from Euclidean distance.

Figures 2–3, p.4 · Main paper

Tables, comparisons and evidence

Separates reimplemented prior methods and CLIP-backbone variants; data-size and label-noise tables expose distinct robustness interventions.

Table 1, p.6; Tables 4–5, p.7 · Main paper

Appendix and reproducibility

Directs the inference schematic to supplement Section 7, dataset/training detail to Sections 8–12, and replication caveats to Section 21; those pointers, not the separate supplement, were inspected.

Sections 3.4 and 4.1, pp.5–6 · Main paper

Clarity and writing conventions

The definition of TPS gives its range and perfect/worst endpoints, but Table 3 reverses the hour-error and TPS direction arrows relative to the definitions and neighboring tables.

Equation 11, p.6; Table 3, p.7 · Main paper

What to reuse—and what not to infer

Transferable lesson: Pair every input contract with a matched baseline and explain cyclic variables using one small worked distance example.

Boundary: Audit metric-direction arrows and numerical prose: Table 3 arrows conflict with metric meaning, and the prose describing the 5% data month-error rise understates the displayed 1.40 to 1.89 change. Qualitative text queries are not a broad retrieval benchmark.

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