- Rating
- 3
- Confidence
- 3
- Informativeness
- 2
- Presentation
- 2
- Soundness
- 2
Read review
Paper summary
GazeEarth targets UHR remote sensing VQA under a fixed input budget. Two probes set it up: a frozen MLLM can usually name the grid cell a question points at, but feeding back that crop costs 9.2 points on MME-RealWorld-RS against a plain resized overview. Presentation, not localisation, is the claimed bottleneck. The model picks up to two cells of a 4x4 overview, a separable piecewise linear warp enlarges their enclosing rectangle inside a 2048 pixel canvas while compressing the periphery, and the same model answers from that view. Training free, two calls, three benchmarks, four backbones.
Critical strengths
- Table 7 is the useful experiment: one region, four renderings, same answerer.
- Cheap, training free, unchanged across four backbones, and the statistics are done properly.
Critical weaknesses
- Backbones differ throughout Tables 3 and 4 (ZoomEarth 3B, most RS models 7B, GazeEarth on Qwen3-VL-8B), so beating task trained methods proves less than claimed. Against Overview on the same model the gains are +8.38, +3.42 and +1.01, the last not surviving correction, and 4.07 of the 8.34 point mean gain over Direct is canvas normalisation, which is plain resizing.
- With δ = 0.25, α = 1.5 and β = 0.55 the focus grows about 1.5x linearly over a uniform resize, taking a 20 pixel object on a 10k scene to roughly 6 pixels. Thin basis for 8 to 10 points, and counting agrees: 21.78 on MME-RealWorld-RS, up 0.43 over Direct, behind ZoomEarth at 31.66. Stratify by target size, and report how often a request ends as global_only or brief_incomplete rather than a real focus. Two distant cells also give an enclosing rectangle covering most of the scene, so the case that motivates K = 2 gets almost no enlargement.
- fx and fy are applied independently, so two cross shaped bands come out stretched on one axis and squeezed on the other, and aspect ratio is a category cue for runways and ships. The prompt never says the canvas is geometrically non uniform, yet many questions ask about position on it. Section 3.3 overstates things: a product of two monotone maps is a homeomorphism by construction, and Eq. 6 is an identity.
- Two-Gaze and Three-Gaze reselect on an already warped canvas and warp it again, so later rounds undo earlier ones and nothing accumulates. Degradation is what that design predicts; the XLRS-Bench fall from 53.51 to 46.27 reads as composition damage, not evidence that one fixation suffices. A fair counterpart would carry information across fixations and stop when it has enough. AdaptiveNN (Wang et al., Emulating human like adaptive vision for efficient and flexible machine visual perception, Nature Machine Intelligence 2025) does that, with sequential fixations, incremental aggregation and an explicit stopping rule under an adjustable budget. Even a paragraph of discussion would support the single gaze claim better than the present baseline.
- The 4x4 grid, K = 2, α = 1.5 and the 1536 pixel selector are each the argmax of the mean over the evaluation sets, the sweeps are not monotone (XLRS-Bench: 50.97, 53.51, 50.49), and the spread rivals the margin over the baselines. No held out split. The 77.74% alignment is partly circular too, since reference cells come from spatial phrases in the question itself.
Questions for authors
See Weakness
11 comments
Response to Reviewer svEJ Part (2/8)
Table S3. Answer accuracy by target size in the source image.
| Benchmark | Native target size | Questions | Overview acc. (%) | GazeEarth acc. (%) | Gain (pp) |
|---|---|---|---|---|---|
| XLRS-Bench | Small (<32 px) | 323 | 35.91 | 43.96 | +8.05 |
| XLRS-Bench | Medium (32–<96 px) | 709 | 37.38 | 46.54 | +9.17 |
| XLRS-Bench | Large (≥96 px) | 628 | 57.01 | 65.13 | +8.12 |
| XLRS-Bench | All coordinate-resolved questions | 1,660 | 44.52 | 53.07 | +8.55 |
| XLRS-Bench | ↳ Tiny (<16 px), included in Small | 93 | 27.96 | 36.56 | +8.60 |
| LRS-GRO | Small (<32 px) | 31 | 67.74 | 64.52 | −3.23 |
| LRS-GRO | Medium (32–<96 px) | 865 | 57.69 | 59.19 | +1.50 |
| LRS-GRO | Large (≥96 px) | 6,498 | 53.89 | 55.19 | +1.29 |
| LRS-GRO | ↳ Tiny (<16 px), included in Small | 0 | — | — | — |
On the coordinate-resolved XLRS-Bench subset, gains are comparable across Small, Medium, and Large targets. On LRS-GRO, GazeEarth improves on Medium and Large targets. The Small group contains only 31 questions: GazeEarth answers 20 correctly and Overview answers 21.
Your 20-pixel example identifies a real resolution limit. Resizing a 10,000-pixel image axis to 2,048 pixels reduces a 20-pixel target to about 4.1 pixels along that axis. Even with the full 1.5× axial enlargement, it spans only about 6.1 pixels. On the 93 XLRS Tiny questions, GazeEarth answers 34 correctly (36.56%), compared with 26 (27.96%) for Overview. This recovers eight additional answers, but absolute accuracy remains low. These questions represent 5.60% of the coordinate-resolved subset and 3.02% of the full XLRS benchmark. We found no Tiny targets in the annotated LRS-GRO subset; MME lacks the boxes needed to estimate their prevalence. Meanwhile, Table 11 in our article also verifies this point. In this table, we can see that Self-MLLM still achieved the best on 537 questions without specific direction evidence from LRS-GRO. Among them, Center hit reached 64.99%, Box coverage reached 55.65, which is significantly better than other methods, and the answer accuracy rate of Oracle Gaze is similar (Self- MLLM: 66.29, Oracle Gaze: 70.20).
Exhaustive counting exposes a related coverage limitation. On the 1,226 MME counting questions, GazeEarth answers 267 correctly (21.78%), compared with 225 (18.35%) for Overview. Trained ZoomEarth remains stronger on this category, as you note. When instances are scattered across the image, a single focus enlarges one neighborhood while other instances remain at overview resolution. Trained methods with multi-step zooming can cover more instances; limited coverage is a cost of our single-gaze, training-free design. The selector also declares brief_incomplete more often on counting questions than on other MME questions (15.25% versus 5.53%). We will discuss this limitation and identify counting-oriented scanning as a direction for future work.
Response to Reviewer svEJ Part (3/8)
(2) Request outcomes and actual enlargement. We counted global_only and brief_incomplete requests, their proportions, and their conditional answer accuracies across every question in all three benchmarks. The statistics also include valid selections, invalid outputs, and XLRS-Bench requests whose focus is determined by question coordinates. We separately inspected the final rendered views to identify all cases without actual enlargement.
Table S4. Request outcomes and conditional answer accuracy across all three benchmarks.
| Request outcome | XLRS-Bench, n (%) | Accuracy (%) | MME-RealWorld-RS, n (%) | Accuracy (%) | LRS-GRO, n (%) | Accuracy (%) |
|---|---|---|---|---|---|---|
| Coordinate-derived focus | 1,660 (53.90) | 53.07 | 0 (0.00) | — | 0 (0.00) | — |
| Valid model selection | 1,193 (38.73) | 54.40 | 3,295 (88.15) | 56.33 | 8,884 (91.27) | 59.95 |
| global_only | 90 (2.92) | 46.67 | 54 (1.44) | 40.74 | 489 (5.02) | 79.75 |
| Declared brief_incomplete | 112 (3.64) | 52.68 | 326 (8.72) | 16.56 | 117 (1.20) | 53.85 |
| Invalid response | 25 (0.81) | 68.00 | 63 (1.69) | 44.44 | 244 (2.51) | 65.16 |
| Full benchmark | 3,080 (100.00) | 53.51 | 3,738 (100.00) | 52.43 | 9,734 (100.00) | 61.00 |
| global_only or declared brief_incomplete | 202 (6.56) | 50.00 | 380 (10.17) | 20.00 | 606 (6.23) | 74.75 |
| No realized enlargement | 349 (11.33) | 52.44 | 477 (12.76) | 24.53 | 1,498 (15.39) | 75.43 |
The first five rows are mutually exclusive request outcomes, and the last two rows are summaries. Every proportion uses the full benchmark as its denominator; accuracy is conditional on the questions in that row. The no-enlargement summary also includes invalid responses and valid selections that produced no enlargement. Request status and realized enlargement are therefore different quantities.
Only XLRS-Bench contains questions that trigger coordinate-derived focus from coordinates supplied in the prompt. The compared baselines use the same coordinate-handling policy as GazeEarth. Table S5 reports the control that forces grid selection on every XLRS-Bench question.
Table S5. Effect of forcing grid selection on all XLRS questions.
| XLRS evaluation group | Questions | Coordinate policy (%) | Forced grid selection (%) | Paired difference (pp) |
|---|---|---|---|---|
| Full benchmark | 3080 | 53.51 | 50.13 | −3.38 |
| Questions with resolved coordinates | 1660 | 53.07 | 46.81 | −6.26 |
| Questions without resolved coordinates | 1420 | 54.01 | 54.01 | 0.00 |
Forcing grid selection lowers accuracy on coordinate-bearing questions and leaves the other questions unchanged. We will report this measurable contribution separately from autonomous gaze selection.
Bypass and conditional routing are common implementation practices in related UHR work. GazeEarth's coordinate bypass is one instance of this practice. Public implementations of ZoomEarth, GeoLens, and WeaveEarth also provide such options; the exact files and line numbers are listed below.
Response to Reviewer svEJ Part (1/8)
Thank you for recognizing our informative ablations, rigorous statistical evaluation, and efficient, training-free design. We address your concerns in their original order below.
W1 · Backbones, significance, and gain baselines
(1) Backbone comparisons. Thank you for your constructive advices. To address the backbone difference, we added experiments on Qwen2.5-VL-3B, used by ZoomEarth, and Qwen2.5-VL-7B, used by GeoLens. GazeEarth remains training-free on both and achieves competitive performance on the same backbones as these trained methods.
Table S1. Direct and GazeEarth on the same 3B and 7B backbones; accuracy (%).
GazeEarth improves over Direct on all three benchmarks, with mean gains of 8.70 percentage points (pp) on Qwen2.5-VL-3B and 6.44 pp on Qwen2.5-VL-7B. We then reproduced the training of ZoomEarth and GeoLens on Qwen3-VL-8B to compare the methods on the same newer backbone.
Table S2. Trained methods and GazeEarth on Qwen3-VL-8B; accuracy (%).
In Table S2,
—denotes an unreported mean. The comparison on the same backbone shows that training-free GazeEarth scores higher than both retrained methods on all three benchmarks. The retraining results also indicate that the original training data and recipes do not transfer readily to Qwen3-VL-8B. Some results even fall below Direct: on MME-RealWorld-RS, retrained ZoomEarth scores 31.25%, compared with 42.11% for Direct on the same backbone. As discussed in Section 2, current MLLMs already have strong visual capabilities. Together, the retraining results and GazeEarth's training-free gains support our central argument in Section 2. How visual evidence is presented may therefore matter more for UHR understanding than task-specific training. We agree that the benefit of our presentation mechanism depends on the vision encoder's perceptual capacity and will discuss this dependence. Nevertheless, the consistent improvements across the frozen 3B, 7B, and 8B backbones support the broader applicability of GazeEarth.(2) Statistical significance. As you note, the LRS-GRO gain over Overview does not remain statistically significant after Holm correction (Table 21 in the paper). We agree and have revised the corresponding statement. The gains on XLRS-Bench and MME-RealWorld-RS remain significant under the same test. Thank you again for your detailed discovery.
(3) Ablation baselines. Direct serves as a raw-input baseline, with no image processing by our pipeline, including resizing. Overview applies uniform canvas normalization. We agree that the gain descriptions should distinguish normalization from the additional benefit of directed focus, and have revised them accordingly.
W2 · Magnification, target size, counting, and request status
(1) Magnification, target size, and counting. We agree that the nominal focus strength alone cannot explain the gain over Direct, which also includes canvas normalization. The comparison with Overview measures the additional benefit of directed focus. We have revised the main text to distinguish these contributions.
We stratified questions by target size in the source image, before resizing or warping. For a target box of width w and height h, we define its size as s=wh. Following the COCO thresholds, Small uses s<32 px, Medium uses 32≤s<96 px, and Large uses s≥96 px. Tiny (s<16 px) is reported separately as a subset of Small. For XLRS-Bench, Table S3 uses the same 1,660 coordinate-resolved questions and original-pipeline prediction file as the coordinate-subset results in Tables S4 and S5. These questions account for 53.90% of the benchmark. On LRS-GRO, 7,394 questions (75.96%) have valid annotated object or region boxes. MME-RealWorld-RS does not provide target boxes for this analysis.