Thesis supervisor: Levente Hajder
co-supervisor: Imre Benedek Juhász
Location of studies (in Hungarian): ELTE IK Abbreviation of location of studies: ELTE
Description of the research topic:
The development of the functions of a self-driving car is currently taking place in several areas (sensor perception, fusion behavior planning, intervention, data enrichment, etc.) in parallel. In order to achieve the highest level of autonomous driving, it is important to determine the accuracy and precision of the given self-driving function. A straightforward way to achieve the goal is to look at the process, the “result” of self-driving car, from an external source. With the help of aerial cameras from external sources, the self-driving process and its accuracy can be observed, for which public data and industry data are now available.
Advanced driver assistance systems (ADAS) must be developed on the basis of highly reliable reference data (so-called ground truth) that accurately represent the physical reality and faithfully represent the environment in which the vehicle is moving. The reference data generated in this way can be used to tune, test, and validate the self-driving function. In addition, such data is essential for creating machine-learning-based solutions.
As the currently applied solutions from the earth surface do not meet the expectations described above in all respects, the dissertation focuses on a novel approach. The reference data is generated from air, either by means of sensors mounted on one or more drones or on infrastructure elements. The aim of the dissertation is to explore the different tasks (use cases) and to develop a framework along them, which create an accurate representation of objects by air. Part of the task is to increase the accuracy of the forecast / object representation and to define the necessary framework for the process.
Required language skills: English Recommended language skills (in Hungarian): Hungarian Further requirements: SW: python (Numpy, Pandas, Matplotlib), machine learning: TensorFlow, SKLearn, Pytorch, Keras, optional system development: Spark
Proactivity, good communication skills, enthusiasm about data, cooperation ability
Number of students who can be accepted: 1
Deadline for application: 2024-05-31
2024. IV. 17. ODT ülés Az ODT következő ülésére 2024. június 14-én, pénteken 10.00 órakor kerül sor a Semmelweis Egyetem Szenátusi termében (Bp. Üllői út 26. I. emelet).
Login
