MTMMC
收藏资源简介:
MTMMC数据集是由韩国科学技术院创建的一个大规模真实世界多模态相机跟踪基准。该数据集包含由16个多模态相机在校园和工厂环境中捕捉的长时间视频序列,涵盖了不同时间、天气和季节条件,为研究多相机跟踪提供了挑战性的测试平台。数据集不仅包括RGB图像,还增加了空间对齐和时间同步的热像数据,以提高多相机跟踪的准确性。MTMMC数据集的应用领域包括视觉监控、人群行为分析和异常检测,旨在解决现有数据集在模拟或控制环境下的局限性,推动多目标多相机跟踪技术的发展。
The MTMMC dataset is a large-scale real-world multi-modal camera tracking benchmark developed by the Korea Advanced Institute of Science and Technology (KAIST). It contains long-duration video sequences captured by 16 multi-modal cameras in campus and factory environments, covering diverse temporal, weather and seasonal conditions, thus providing a challenging testbed for multi-camera tracking research. The dataset not only includes RGB images, but also adds spatially aligned and temporally synchronized thermal imaging data to improve the accuracy of multi-camera tracking. The application fields of the MTMMC dataset cover visual surveillance, crowd behavior analysis and anomaly detection. It aims to address the limitations of existing datasets that are restricted to simulated or controlled environments, and promote the advancement of multi-object multi-camera tracking technologies.

- 1MTMMC: A Large-Scale Real-World Multi-Modal Camera Tracking Benchmark韩国科学技术院 · 2024年



