# Aerial image reasoning on HRVQA-VL 1.0 (test) Source: [Aerial image reasoning on HRVQA-VL 1.0 (test)]() Collection index: [llms.txt]() Data updated: 2026-05-07T08:09:06.149Z Evaluation of model performance on the HRVQA-VL 1.0 test set for aerial image reasoning tasks. A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage. | Scope | Recorded value | | --- | --- | | Task | Aerial image reasoning | | Dataset (including any stated split) | HRVQA-VL 1.0 (test) | | Metric | Stage A Accuracy | | Unit | PERCENT | | Direction | Higher is better | | Results loaded | 9 of 9 | | Coverage | All results currently returned for this benchmark | Evaluation settings may differ between results. The original papers and recorded configurations are needed to establish comparability. A shared metric name alone does not establish an identical protocol. ## Metrics - [Stage A Accuracy](): Higher is better; unit: PERCENT. - [Stage B Accuracy](): Higher is better; unit: PERCENT. - [Stage C Accuracy](): Higher is better; unit: PERCENT. - [Overall Accuracy](): Higher is better; unit: PERCENT. ## Recorded results | Method | Stage A Accuracy | Result date | Report paper | | --- | --- | --- | --- | | UAV-VL-R1 | 86.38 | Not recorded | 68fa7ab11e209131cea84a92 | | UAV-VL-R1 | 86.27 | Not recorded | 68fa7ab11e209131cea84a92 | | UAV-VL-R1 | 64.01 | Not recorded | 68fa7ab11e209131cea84a92 | | UAV-VL-R1 | 61.66 | Not recorded | 68fa7ab11e209131cea84a92 | | Qwen2-VL-72B-Instruct | 56.82 | Not recorded | 68fa7ab11e209131cea84a92 | | Qwen2-VL-7B-Instruct | 49.95 | Not recorded | 68fa7ab11e209131cea84a92 | | InstructBLIP-7B | 38.77 | Not recorded | 68fa7ab11e209131cea84a92 | | Qwen2-VL-2B-Instruct | 35.82 | Not recorded | 68fa7ab11e209131cea84a92 | | Qwen2-VL-2B-Instruct | 35.16 | Not recorded | 68fa7ab11e209131cea84a92 | ### UAV-VL-R1 Stage A Accuracy: 86.38. Result ID: 69fc489747f34ce924dedd2c. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Training Stage":"Stage A only","Evaluation Mode":"Zero-shot","Prompting":"Structured"}. This record does not assert independent reproduction. ### UAV-VL-R1 Stage A Accuracy: 86.27. Result ID: 69fc489747f34ce924dedd2f. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Training Stage":"All stages","Evaluation Mode":"Zero-shot","Prompting":"Structured"}. This record does not assert independent reproduction. ### UAV-VL-R1 Stage A Accuracy: 64.01. Result ID: 69fc489747f34ce924dedd2d. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Training Stage":"Stage B only","Evaluation Mode":"Zero-shot","Prompting":"Structured"}. This record does not assert independent reproduction. ### UAV-VL-R1 Stage A Accuracy: 61.66. Result ID: 69fc489747f34ce924dedd2e. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Training Stage":"Stage C only","Evaluation Mode":"Zero-shot","Prompting":"Structured"}. This record does not assert independent reproduction. ### Qwen2-VL-72B-Instruct Stage A Accuracy: 56.82. Result ID: 69fc489747f34ce924dedd2a. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Evaluation Mode":"Zero-shot","Prompting":"Structured"}. This record does not assert independent reproduction. ### Qwen2-VL-7B-Instruct Stage A Accuracy: 49.95. Result ID: 69fc489747f34ce924dedd29. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Evaluation Mode":"Zero-shot","Prompting":"Structured"}. This record does not assert independent reproduction. ### InstructBLIP-7B Stage A Accuracy: 38.77. Result ID: 69fc489747f34ce924dedd28. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Evaluation Mode":"Zero-shot","Prompting":"Structured"}. This record does not assert independent reproduction. ### Qwen2-VL-2B-Instruct Stage A Accuracy: 35.82. Result ID: 69fc489747f34ce924dedd2b. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Version":"SFT","Evaluation Mode":"Zero-shot","Prompting":"Structured"}. This record does not assert independent reproduction. ### Qwen2-VL-2B-Instruct Stage A Accuracy: 35.16. Result ID: 69fc489747f34ce924dedd27. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Version":"Base","Evaluation Mode":"Zero-shot","Prompting":"Structured"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.