Strojové učení pro řízení simulovaných vozidel
Thesis title in Czech: | Strojové učení pro řízení simulovaných vozidel |
---|---|
Thesis title in English: | Machine Learning for Driving of Virtual Vehicles |
Key words: | Umělá Inteligence, Strojové Učení, Navigace, Simulace |
English key words: | Artificial Intelligence, Machine Learning, Navigation, Simulation |
Academic year of topic announcement: | 2018/2019 |
Thesis type: | diploma thesis |
Thesis language: | čeština |
Department: | Department of Software and Computer Science Education (32-KSVI) |
Supervisor: | Mgr. Jakub Gemrot, Ph.D. |
Author: | hidden - assigned and confirmed by the Study Dept. |
Date of registration: | 08.01.2019 |
Date of assignment: | 08.01.2019 |
Confirmed by Study dept. on: | 17.01.2019 |
Date and time of defence: | 14.09.2020 08:30 |
Date of electronic submission: | 30.07.2020 |
Date of submission of printed version: | 30.07.2020 |
Date of proceeded defence: | 14.09.2020 |
Opponents: | Mgr. Vladan Majerech, Dr. |
Guidelines |
Cars in virtual worlds such as games are typically driven by hand coded algorithms with hand tuned parameters. That is laborious and it needs to be updated each time simulation changes or new car types are added.
The aim of this thesis is to apply machine learning and artificial intelligence techniques to teach physically simulated cars how to drive in a 3D virtual world. The resulting driving model should directly control vehicle steering, throttle and brake. The car should be able to follow a route on road. Results should be presented visually. |
References |
Wierstra, D., Schaul, T., Glasmachers, T., Sun, Y., Peters, J., Schmidhuber, J. (2014). Natural evolution strategies. Journal of Machine Learning Research, 15(1), 949-980. Retrieved from http://www.jmlr.org/papers/volume15/wierstra14a/wierstra14a.pdf
Schulman, J., Wolski, F., Dhariwal, P., Radford, A., Klimov, O. (2017). Proximal Policy Optimization Algorithms. Retrieved from: https://arxiv.org/abs/1707.06347 Salimans, T., Ho, J., Chen, X. Sutskever, I. (2017). Evolution Strategies as a Scalable Alternative to Reinforcement Learning. Retrieved from https://arxiv.org/pdf/1703.03864.pdf Ha, D., Schmidhuber, J. (2018). World Models. Retrieved from: https://arxiv.org/abs/1803.10122 |