Dataset for Enhancing Camera-based Indoor Positioning with Neural Networks

This dataset was created for the purposes of a master's thesis focused on vision-based indoor positioning through video classification using neural networks. It contains smartphone-captured videos of empty indoor corridors at the University of Pavol Jozef Šafárik (Jesenná 5 and Park Angelinum buildings) in Košice, Slovakia. Each video consists of 50 frames and the dataset contains both original and augmented data to enrich the training set. The dataset is organized into 9 classes, each corresponding to a distinct but spatially connected corridor segment. The spatial layout and relationships between these segments are illustrated in the accompanying file floorplan_with_segments.png. Note: In line with the thesis' focus on privacy and anonymisation, only videos featuring empty corridors are publicly available. Videos containing individuals were deliberately excluded in order to maintain privacy standards.