Sirvi Autor "Zabolotnii, Dmytro, juhendaja" järgi

Nüüd näidatakse 1 - 2 2

Machine Learning Solutions for the Task of Pedestrian Trajectory Prediction – A Systematic Literature Review
(Tartu Ülikool, 2024) Garg, Ankit; Muhammad, Naveed, juhendaja; Zabolotnii, Dmytro, juhendaja; Tartu Ülikool. Loodus- ja täppisteaduste valdkond; Tartu Ülikool. Arvutiteaduse instituut
This study aims to provide an in-depth overview of existing methodologies, trends, and challenges in human trajectory prediction. It analyzes diverse literature to examine various approaches involved in machine learning techniques. This study categorizes these methodologies based on their foundational principles, delving into their strengths and limitations. Particular emphasis is placed on recent advances in machine learning mixed with psychological and environmental aspects for human trajectory prediction. This study finds three significant categories: cognitive approaches, pattern-based approaches, and probabilistic approaches. These are then further divided into different sub-categories, thus forming a taxonomy. Categories at each level of the hierarchical taxonomy are compared, with information about their pros, cons, and where each category should be used. Furthermore, the research papers studied during this survey were split into categories based on their methods. In conclusion, it was found that the “Behavioral Features Method” category performed the best among the other categories. Thus, more research should be done on combining machine learning methods with behavioral features.
Vision-based Localization on City Scale Using Open Street Map
(Tartu Ülikool, 2025) Sokk, Helena; Muhammad, Naveed, juhendaja; Zabolotnii, Dmytro, juhendaja; Tartu Ülikool. Loodus- ja täppisteaduste valdkond; Tartu Ülikool. Arvutiteaduse instituut
Autonomous vehicles, like other robots, need to localize themselves in order to navigate. While global navigation satellite systems (GNSS) such as GPS can provide such vehicles with localization information, the GNSS information might not always be available. Since localization is one of the crucial components in self-driving vehicles, it is important to develop robust techniques to accomplish it. One such localization technique for vehicles to localize is using particle filters, given a map of the environment. The goal of this thesis was to implement a robust localization framework that integrates the particle filter with vision-based street sign detection and Open Street Map, without any reliance on GNSS. The proposed framework was able to localize the vehicle within a radius of 10 meters of its ground truth location, showing promising results. The implemented framework provides a good starting point for any future improvements and experiments in the problem of GNSS-free localization in autonomous vehicles.