MSLS

该数据集是为了基准测试基于序列的位置识别而新发布的，它包含了来自不同城市在各种外观条件下的图像序列。在训练阶段，该数据集使用了墨尔本的数据，而在具有不同条件的其他城市进行测试。该任务旨在实现视觉位置识别。

This newly released dataset is designed for benchmarking sequence-based place recognition. It comprises image sequences collected from various cities under diverse visual appearance conditions. During the training phase, data from Melbourne is utilized, while testing is conducted in other cities with distinct conditions. The task aims to achieve visual place recognition.