Amirkabir campus dataset (AUT-VI)

Amirkabir校园数据集（AUT-VI）是由伊朗德黑兰的阿米尔卡比尔理工大学开发的一个新型且极具挑战性的数据集，旨在解决视觉惯性里程计（VIO）在动态环境中的应用问题，特别是在为视障人士提供室内外导航服务时。该数据集包含126个多样化的序列，分布在17个不同的位置，涵盖了动态对象、挑战性的环路闭合/地图重用、不同的光照条件、反射以及突然的相机移动等多种极端导航场景。数据集的创建过程涉及使用Android应用程序进行数据捕获，并公开发布以支持研究社区的进一步开发。AUT-VI数据集主要应用于评估和改进为视障人士设计的导航工具，特别是在动态和复杂环境中提高定位算法的准确性和鲁棒性。

The Amirkabir Campus Dataset (AUT-VI) is a novel and challenging dataset developed by Amirkabir University of Technology in Tehran, Iran. It aims to address the application of Visual-Inertial Odometry (VIO) in dynamic environments, especially for indoor and outdoor navigation services for visually impaired individuals. This dataset includes 126 diverse sequences distributed across 17 distinct locations, covering a variety of extreme navigation scenarios such as dynamic objects, challenging loop closure/map reuse, varying lighting conditions, reflections, and sudden camera movements. The dataset was created using an Android application for data capture, and is publicly released to support further development within the research community. The AUT-VI dataset is primarily used to evaluate and improve navigation tools designed for visually impaired individuals, particularly to enhance the accuracy and robustness of positioning algorithms in dynamic and complex environments.