Computer Vision & Pattern Recognition
Outdoor positioning systems based on the Global Navigation Satellite System have several shortcomings that have deemed their use for positioning inside buildings and in urban canyons impractical. Location fingerprinting, which utilizes machine learning, has emerged as a viable solution for such scenarios due to its simple concept and good results. Shallow learning algorithms are traditionally used in location fingerprinting.
Outdoor positioning systems based on the Global Navigation Satellite System have several shortcomings that have deemed their use for positioning inside buildings and in urban canyons impractical. Location fingerprinting, which utilizes machine learning, has emerged as a viable solution for such scenarios due to its simple concept and good results. Shallow learning algorithms are traditionally used in location fingerprinting. Recently, however, the research community started utilizing deep learning for fingerprinting after witnessing the great success and superiority these methods have over shallow machine learning. Thus, this project aims to develop deep learning-based solutions that can push the boundaries of the filed. These include, for example, utilizing deep learning to enhance positioning accuracy, provide robust performance against measuring noise, and reduce site surveying costs. Additionally, due to the lack of publicly available datasets that researchers can use to develop, evaluate, and compare fingerprint-based positioning solutions, another goal of this project is to collect and make publicly available real-world data, such as WiFi, Bluetooth, Cellular, and sensor measurements, that are annotated with reliable ground truth coordinates.
