RGB-Thermal MOT dataset

该数据集是一组在多样化城市环境中捕获的大型热成像和RGB图像集合，这些图像针对多目标检测和追踪进行了标注。它是世界上首个针对多目标追踪（MOT）的大规模RGB及其对应热成像数据集，专注于行人检测和多目标追踪研究。该数据集包含30个序列，每种模态总计9000帧，旨在支持多目标追踪任务。This dataset is a large collection of thermal imaging and RGB images captured in diverse urban environments, with annotations prepared for multi-object detection and tracking. It is the world's first large-scale RGB and corresponding thermal imaging dataset tailored for multi-object tracking (MOT), focusing on pedestrian detection and multi-object tracking research. This dataset includes 30 sequences, with a total of 9000 frames per modality, and is designed to support multi-object tracking tasks.

搜集汇总

背景与挑战

背景概述

该数据集是一个大规模的热成像和RGB图像集合，采集自多样化城市环境，专为多目标追踪任务设计。它作为热成像领域的新基准资源，旨在提升追踪准确性和鲁棒性，并支持相关研究方法的实验验证。