<b><i>Kust4K:</i></b><i> </i><b><i>An RGB-TIR Dataset from UAV Platform for Robust Urban Traffic Scenes Semantic Segmentation</i></b>

We introduce Kust4K, a multimodal semantic segmentation dataset, which is the first and largest UAV-based dataset for RGB-TIR semantic segmentation. Kust4K dataset is designed to overcome key limitations in existing UAV-based semantic segmentation datasets: low information density, limited data volume, and insufficient robustness discussion under non-ideal environment. Kust4K dataset featuring 4,024 of 640×512 pixel-aligned RGB-Thermal Infrared image pairs captured across diverse urban road scenes under variable illumination. Extensive experiments with state-of-the-art models demonstrate Kust4K’s effectiveness, with multimodal training, significantly outperforming unimodal baselines. Additionally, these results highlight that multimodal image information is critical for obtaining more reliable semantic segmentation results. In total, Kust4K dataset advance robust urban traffic scene understanding, offering a valuable resource for intelligent transportation research.本研究提出Kust4K多模态语义分割数据集，它是首个且规模最大的基于无人机（Unmanned Aerial Vehicle, UAV）的可见光-热红外（RGB-Thermal Infrared, RGB-TIR）语义分割数据集。Kust4K数据集旨在克服现有基于无人机的语义分割数据集存在的核心局限：信息密度偏低、数据规模有限，且未充分开展非理想环境下的模型鲁棒性相关研究。该数据集包含4024对分辨率为640×512的像素对齐可见光-热红外图像对，采集场景涵盖多种城市道路环境，且采集时的光照条件存在变化。借助当前最优模型开展的大量实验验证了Kust4K数据集的有效性：采用多模态训练的模型性能显著优于单模态基准模型。此外，实验结果表明，多模态图像信息对于获取更可靠的语义分割结果具有关键作用。总体而言，Kust4K数据集为实现更鲁棒的城市交通场景理解提供了支撑，同时为智能交通领域的研究提供了宝贵的资源。