EarthJev: Efficient Visual Evidence Acquisition and Decision for UHR Remote Sensing Understanding

Pengyu Dai; Yao Zhang; Jian Song; Junjue Wang; Yafei Ou; Yuzeng Chen; Junshi Xia; Hongruixuan Chen

cs.CVNov 4, 2026

EarthJev: Efficient Visual Evidence Acquisition and Decision for UHR Remote Sensing Understanding

Authors: Pengyu Dai, Yao Zhang, Jian Song, Junjue Wang, Yafei Ou, Yuzeng Chen, Junshi Xia, Hongruixuan Chen,

Organizations: RIKEN Center for Advanced Intelligence Project (AIP), RIKEN, Tokyo, Japan. · The University of Tokyo, Japan. · the State Key Laboratory of Information Engineering in Surveying, Mapping and Remote Sensing, Wuhan University, Wuhan, China.

Abstract

Ultra-high-resolution (UHR) remote sensing visual question answering (VQA) requires locating fine-grained evidence in large scenes under limited visual input budgets. Active observation addresses this challenge through selective exploration, but repeated generative judgments can make region search costly. Its decisions also concern distinct prediction targets: spatial pointer validity, evidence presence, and reasoning correctness. We propose EarthJev, a framework that separates region scoring from answer generation and uses target-specific decision signals while keeping the answering model frozen. A spatial pointer verifier uses visual and geometric cues to guide exploration or bypass search. An evidence-aware decision mechanism scores compact candidate views in batches using a lightweight head on frozen representations, without generating textual region judgments. Format-specific confidence calibration then guides selection between candidate answers generated from memory and overview presentations of the acquired evidence. Across five benchmarks, EarthJev achieves a mean accuracy of 53.06% with Qwen3.5-9B, outperforming ZoomSearch with the same answering model by 2.33 to 14.73 percentage points per benchmark while reducing mean latency from 11.30 - 1.67 seconds per question. With four additional frozen answering models, model-specific fitting of the decision modules and calibration mappings yields mean accuracy gains of 3.41 - 6.53 percentage points over the corresponding single-pass baselines. Code is available at https://github.com/Da1daidaidai/EarthJev

Figures & tables

Fig. 1: Challenges in active observation: ambiguous spatial cues, misleading search history, and uncertain answers despite relevant evidence. Cyan marks target-relevant regions and evidence, red indicates distractors and off-target search, and yellow boxes delineate crop regions.

Fig. 2: Preliminary diagnostics of active observation: (a) Spatial cues: valid cues aid perception whereas wrong cues degrade it; (b) Evidence coverage: heuristic search (ZS) and Yes/No probing (YN) frequently miss target evidence; and (c) Answer scoring: answer confidence predicts answer correctness far better than evidence scores.

Fig. 3: Comparison of visual evidence acquisition paradigms. ZoomSearch employs crop-based search, while GazeEarth uses non-uniform resizing. Both may focus on incomplete or misleading evidence, resulting in incorrect answers in this example. In contrast, EarthJev verifies spatial cues, reassesses candidate views during search, and selects among candidate answers using calibrated confidence, enabling more accurate identification.

Fig. 4: The EarthJev workflow consists of three stages: (1) Spatial pointer verification, which integrates visual and geometric information to evaluate the localization cues provided in a question, determining whether the perspective presented by these cues is sufficient to answer the question directly or if further searching is required. (2) An EDM, which utilizes vision-language representations to score candidate regions, thereby guiding hierarchical expansion and the retention of evidence. (3) The selected evidence is organized into a “memory” and a “canvas”, which are then accessed by a shared, frozen answering model. Finally, the system evaluates candidate answers based on calibrated confidence scores and selects the most reliable one as the definitive answer.

Method	Backbone	LRS-VQA				MME-RS				LRS-GRO				RSHR-Bench			XLRS-lite	Overall
		FAIR	Bridge	STAR	Avg.	Color	Count	Position	Avg.	Global	Region	Object	Avg.	Perception	Reasoning	Avg.	Avg.	Avg.
ZoomSearch [ 40 ]	Qwen3.5-9B	29.75	31.26	34.31	32.46	70.20	39.31	71.84	60.62	67.14	48.52	50.31	53.91	35.74	55.97	37.52	44.48	45.80
GazeEarth [ 41 ]	Qwen3-VL-8B	29.18	26.74	29.38	28.94	63.67	21.78	71.12	52.43	77.96	50.39	58.04	61.00	44.49	59.27	46.45	52.95	48.36
ViCrop [ 36 ]	Qwen3-VL-8B	27.33	29.57	30.51	29.39	46.85	16.97	47.02	37.11	75.56	51.48	54.03	58.58	34.58	61.67	36.05	43.51	40.93
AttWarp [ 38 ]	Qwen3-VL-8B	26.76	31.54	30.96	29.74	54.58	19.00	62.13	45.45	77.75	49.00	53.39	58.18	33.80	60.34	36.22	45.36	42.99
WeaveEarth [ 39 ]	Qwen3-VL-8B	27.64	24.95	34.78	31.15	54.02	23.49	65.71	47.94	75.34	47.74	56.50	58.94	32.74	63.76	36.14	41.36	43.10

TABLE I: Comparison of EarthJev with existing methods across five remote sensing VQA benchmarks. Results are reported as accuracy (%). The top three distinct scores per column are highlighted in coral, amber, and blue; best results are boldfaced, and ties share the same color. Benchmark Avg. denotes question-level accuracy, while Overall Avg. is the mean across five benchmarks. RSHR-Bench Perception/Reasoning scores are averaged over task categories.

Panel of Table 6 — TABLE II: Performance comparison of Earth with existing methods on XLRS-Bench-lite, covering 13 fine-grained Level-3 (L3) task categories. Training settings are indicated by ✓ (trained), ✓ ✓ ✓ (answering model frozen, with only lightweight decision modules trained), and ✗ (training-free).

Fig. 5: Comparison of overall accuracy across five remote sensing benchmarks.

Group	EDM	SPV	CAS	MME-RS	LRS-VQA	XLRS-lite	LRS-GRO	RSHR-Bench	Avg.
1	✗	✗	✗	60.62	32.46	44.48	53.91	37.52	45.80
2	✓	✗	✗	62.97	34.52	44.29	55.42	38.99	47.24
3	✓	✓	✗	62.81	35.20	55.00	56.04	54.16	52.64
4	✓	✓	✓	63.16	34.79	55.81	59.27	52.25	53.06

TABLE III: Component ablation of EarthJev with Qwen3.5-9B. Results are reported as accuracy (%). The full configuration is shaded in light coral; the best result in each column is boldfaced.

Fig. 6: Visual memory examples combining a global overview with local views indexed by their source locations. Memory and canvas answers are shown with calibrated confidence in parentheses.

Answerer	Setting	MME-RS	LRS-VQA	XLRS-lite	LRS-GRO	RSHR-Bench	Avg.
Qwen3.5-4B	Base	51.95	29.74	41.95	58.11	37.52	43.85
	+ EarthJev	57.76	30.32	53.96	59.01	50.87	50.38
	Gain (pp)	+5.81	+0.57	+12.01	+0.90	+13.34	+6.53
Qwen3-VL-8B	Base	55.97	32.27	46.10	59.27	36.05	45.93
	+ EarthJev	56.39	34.97	54.38	59.40	49.48	50.92
	Gain (pp)	+0.43	+2.70	+8.28	+0.13	+13.43	+4.99

TABLE IV: Performance of EarthJev across answering models (accuracy, %). Gray and coral cells show the base and EarthJev-augmented results, respectively; blue cells report gains in percentage points.

Selection	MME-RS	LRS-VQA	XLRS-lite	LRS-GRO	RSHR-Bench	Avg.
Memory only	62.92	33.15	55.00	58.53	54.16	52.75
Canvas only	56.18	35.20	53.64	56.04	45.84	49.38
Raw confidence	62.17	34.53	56.01	58.88	51.82	52.68
CAS (EarthJev)	63.16	34.79	55.81	59.27	52.25	53.06

TABLE V: Answer selection over the same retrieved evidence with Qwen3.5-9B (accuracy, %). Raw confidence and CAS choose between the memory and canvas answers by raw or calibrated confidence.

Fig. 7: Design evidence for the three modules (Qwen3.5-9B). (a) EDM vs. other decision sources: mean accuracy over the five benchmarks against seconds per question. (b) SPV vs. blind trust: accuracy change on the LRS-VQA pointer test with a correct or a wrong stated box, relative to no box. (c) Held-out calibration error of each module’s decision score (evidence presence, pointer trust, answer confidence) before and after Platt scaling.

Qwen3.5-9B
Method	s/q	Rel. time	Avg.
Direct	3.30	1.0 $\times$	46.93
ZoomSearch	11.30	3.4 $\times$	45.80
Zoom-Refine	11.34	3.4 $\times$	46.94
EarthJev	1.67	0.5 $\times$	53.06
Qwen3-VL-8B

TABLE VI: Inference cost under the same answering model. s/q: seconds per question on one NVIDIA H100 GPU

Fig. 8: Qualitative comparison on spatial reasoning and object counting. EarthJev selects the correct answer by evaluating calibrated confidence across memory and canvas presentations of the retrieved evidence.

Remote Sensing Visual Question Answering

Remote sensing visual question answering (VQA) studies how AI answers natural-language questions about satellite, aerial, and other Earth-observation imagery.

See all 41 papers

GeoJeV: Unified Spatial Probability for Training-Free Efficient Evidence Acquisition in Ultra-High-Resolution Remote Sensing
Nov 6, 2026Xinyi Niu, Lingfeng Chen, Jiehua Zhang +4Efficient Multimodal Inference Remote Sensing VLMs
College of Electronic Science and Technology, National University of Defense Technology, Changsha 410073, Hunan, China.
The Earth in One Gaze: Training-Free Active Focus for UHR Remote Sensing Understanding
Sep 23, 2026Yao Zhang, Pengyu Dai, Wei Guo +5Remote Sensing Image Understanding Remote Sensing VLMs
Wuhan University
RS-OPSD: Reliable Privileged On-Policy-Self-Distillation for Ultra-High-Resolution Remote Sensing VQA
Sep 29, 2026Chengjie Jiang, Yunqi Zhou, Jiafeng Yan +3Remote Sensing Image Understanding Visual Question Answering
Tsinghua University · Zhejiang University · Central University of Finance and Economics +2

Nov 6, 2026J/K move · Enter open · S save

GeoJeV: Unified Spatial Probability for Training-Free Efficient Evidence Acquisition in Ultra-High-Resolution Remote Sensing

Xinyi Niu, Lingfeng Chen, Jiehua Zhang, Panhe Hu+3

College of Electronic Science and Technology, National University of Defense Technology, Changsha 410073, Hunan, China.

Efficient Multimodal Inference Remote Sensing VLMs Remote Sensing Visual Question Answering

Open paper AskPDF arXiv

Ultra-high-resolution (UHR) remote sensing imagery provides rich spatial details but poses challenges for accurate and efficient multimodal understanding. Existing evidence acquisition methods selectively explore relevant image regions, yet largely rely on generative multimodal large language models (MLLMs) for intermediate search control. However, reinforcement learning from human feedback (RLHF) primarily aligns these models with human-preferred responses rather than explicitly optimizing region-level relevance judgments, while repeated autoregressive decoding introduces additional computational overhead. To address these limitations, we propose GeoJeV, a training-free method that formulates evidence acquisition as spatial probability-driven decision making. At each evaluated region, a frozen JeV model produces a probability distribution over candidate regions. Specifically, Region Rank prioritizes candidate regions, Sparse Branch adaptively determines branching width through sparse probability support, and Depth Gate controls refinement depth based on branch sparsity and distribution concentration. By coordinating search priority, branching width, and refinement depth through a shared distribution, GeoJeV reduces reliance on RLHF-aligned generation and eliminates intermediate decoding. Experiments on LRS-GRO, MME-RealWorld-RS, and LRS-VQA achieve accuracies of 60.59%, 52.79%, and 34.32%, respectively. Across the three benchmarks, GeoJeV improves the average accuracy of Qwen3-VL-8B by +5.32pp with only 1.12 additional seconds per question, while achieving a 7.8-fold inference speedup over the recent method.

Abstract

Remote Sensing Visual Question Answering

Explore similar work

GeoJeV: Unified Spatial Probability for Training-Free Efficient Evidence Acquisition in Ultra-High-Resolution Remote Sensing

The Earth in One Gaze: Training-Free Active Focus for UHR Remote Sensing Understanding

RS-OPSD: Reliable Privileged On-Policy-Self-Distillation for Ultra-High-Resolution Remote Sensing VQA