你的 VLM 准备好飞向天空了吗？面向无人机导航的综合空间智能基准
Is your VLM Sky-Ready? A Comprehensive Spatial Intelligence Benchmark for UAV Navigation
为视觉研究者提供了一个面向无人机视角的空间智能评测基准（13 个子类，覆盖包围框、距离、高度、降落安全等），并给出可复用的 SpatialSky-Dataset 与 Sky-VLM 训练范式，可用于诊断和提升 VLM 的细粒度空间推理能力。
原文依据 we introduce SpatialSky-Bench, a comprehensive benchmark specifically designed to evaluate the spatial intelligence capabilities of VLMs in UAV navigation
摘要译文
由模型翻译 · 术语统一 · 请对照原文核对关键结论视觉语言模型（VLM）凭借其强大的视觉感知与推理能力，已被广泛应用于无人机（UAV）任务。然而，现有 VLM 在无人机场景中的空间智能能力仍基本未被探索，这引发了人们对其在动态环境中导航与理解的有效性的担忧。为填补这一空白，我们提出 SpatialSky-Bench，一个专门用于评估 VLM 在无人机导航中空间智能能力的综合基准。该基准包含环境感知与场景理解两个类别，并细分为 13 个子类，涵盖包围框、颜色、距离、高度与降落安全分析等。对多种主流开源与闭源 VLM 的大量评估表明，它们在复杂无人机导航场景中的表现并不理想，凸显出其空间能力上的显著差距。为应对这一挑战，我们构建了 SpatialSky-Dataset，这是一个包含 100 万样本、覆盖多种场景与多样标注的综合数据集。基于该数据集，我们提出 Sky-VLM，一个专为跨多种粒度与上下文进行无人机空间推理而设计的专用 VLM。大量实验结果表明，Sky-VLM 在所有基准任务上均取得最优性能，为开发适用于无人机场景的 VLM 铺平了道路。源代码见 https://github.com/linglingxiansen/SpatialSKy 。
英文原文摘要 1548 字符
Vision-Language Models (VLMs), leveraging their powerful visual perception and reasoning capabilities, have been widely applied in Unmanned Aerial Vehicle (UAV) tasks.However, the spatial intelligence capabilities of existing VLMs in UAV scenarios remain largely unexplored, raising concerns about their effectiveness in navigating and interpreting dynamic environments.To bridge this gap, we introduce SpatialSky-Bench, a comprehensive benchmark specifically designed to evaluate the spatial intelligence capabilities of VLMs in UAV navigation. Our benchmark comprises two categories--Environmental Perception and Scene Understanding--divided into 13 subcategories, including bounding boxes, color, distance, height, and landing safety analysis, among others.Extensive evaluations of various mainstream open-source and closed-source VLMs reveal unsatisfactory performance in complex UAV navigation scenarios, highlighting significant gaps in their spatial capabilities.To address this challenge, we developed the SpatialSky-Dataset, a comprehensive dataset containing 1 M samples with diverse annotations across various scenarios. Leveraging this dataset, we introduce Sky-VLM, a specialized VLM designed for UAV spatial reasoning across multiple granularities and contexts.Extensive experimental results demonstrate that Sky-VLM achieves state-of-the-art performance across all benchmark tasks, paving the way for the development of VLMs suitable for drone scenarios.The source code is available at https://github.com/linglingxiansen/SpatialSKy .