TY - JOUR AU - Bencsik, Blanka AU - Reményi, István AU - Szemenyei, Márton AU - Botzheim, János TI - Designing an Embedded Feature Selection Algorithm for a Drowsiness Detector Model Based on Electroencephalogram Data JF - SENSORS J2 - SENSORS-BASEL VL - 23 PY - 2023 IS - 4 SP - 1874 SN - 1424-8220 DO - 10.3390/s23041874 UR - https://m2.mtmt.hu/api/publication/33629587 ID - 33629587 AB - Driver fatigue reduces the safety of traditional driving and limits the widespread adoption of self-driving cars; hence, the monitoring and early detection of drivers’ drowsiness plays a key role in driving automation. When representing the drowsiness indicators as large feature vectors, fitting a machine learning model to the problem becomes challenging, and the problem’s perspicuity decreases, making dimensionality reduction crucial in practice. For this reason, we propose an embedded feature selection algorithm that can be later utilized as a building block in the system development of a neural network-based drowsiness detector. We have adopted a technique: a so-called Feature Prune Layer is placed in front of the first layer in the architecture; as a result, its weights change regarding the importance of the corresponding input features and are deleted iteratively until the desired number is reached. We test the algorithm on EEG data, as it is one of the best indicators of drowsiness based on the literature. The proposed FS algorithm is able to reduce the original feature set by 95% with only 1% degradation in precision, while the precision increases by 1.5% and 2.7% respectively when selecting the top 10% and top 20% of the initial features. Moreover, the proposed method outperforms the widely popular Principal Component Analysis and the Chi-squared test when reducing the original feature set by 95%: it achieves 24.3% and 3.2% higher precision respectively. LA - English DB - MTMT ER - TY - CHAP AU - Szántó, Mátyás AU - Szemenyei, Márton TI - Reinforcement Learning-based Occlusion Avoidance using RGB-D Images T2 - Proceedings of the Workshop on the Advances of Information Technology 2023 PB - BME Irányítástechnika és Informatika Tanszék CY - Budapest SN - 9789634218968 PY - 2023 SP - 92 EP - 101 PG - 10 UR - https://m2.mtmt.hu/api/publication/33624051 ID - 33624051 LA - English DB - MTMT ER - TY - CHAP AU - Skribanek, Solt AU - Szemenyei, Márton AU - Moni, Róbert ED - Zurada, Jacek M. ED - Tadeusiewicz, Ryszard ED - Pedrycz, Witold ED - Korytkowski, Marcin ED - Scherer, Rafał ED - Rutkowski, Leszek TI - Semantically Consistent Sim-to-Real Image Translation with Neural Networks T2 - Artificial Intelligence and Soft Computing PB - Springer International Publishing CY - Cham SN - 9783031234804 T3 - Lecture Notes in Computer Science, ISSN 0302-9743 ; 13589. PY - 2023 SP - 68 EP - 79 PG - 12 DO - 10.1007/978-3-031-23480-4_6 UR - https://m2.mtmt.hu/api/publication/33585734 ID - 33585734 LA - English DB - MTMT ER - TY - CHAP AU - Szántó, Mátyás AU - Szemenyei, Márton TI - Self-Supervised Occlusion Detection and Avoidance using Differentiable Rendering T2 - 2022 International Symposium on Measurement and Control in Robotics (ISMCR) PB - IEEE SN - 9781665454964 PY - 2022 SP - 1 EP - 8 PG - 8 DO - 10.1109/ISMCR56534.2022.9950574 UR - https://m2.mtmt.hu/api/publication/33267518 ID - 33267518 LA - English DB - MTMT ER - TY - CHAP AU - Bencsik, Blanka AU - Szemenyei, Márton TI - Efficient Neural Network Pruning Using Model-Based Reinforcement Learning T2 - 2022 International Symposium on Measurement and Control in Robotics (ISMCR) PB - IEEE SN - 9781665454964 PY - 2022 SP - 1 EP - 8 PG - 8 DO - 10.1109/ISMCR56534.2022.9950598 UR - https://m2.mtmt.hu/api/publication/33267510 ID - 33267510 LA - English DB - MTMT ER - TY - CONF AU - Princz-Jakovics, Tibor AU - Szemenyei, Márton TI - Environmental Policy Based on the ICT Platform T2 - Conference proceedings of the 3rd International Conference on the Economics of the Decoupling PB - Croatian Academy of Sciences and Arts C1 - Zágráb T3 - Conference proceedings of the International Conference on the Economics of the Decoupling, ISSN 2718-3092 ; 3. PY - 2022 SP - 345 EP - 361 PG - 17 UR - https://m2.mtmt.hu/api/publication/32905384 ID - 32905384 LA - English DB - MTMT ER - TY - CHAP AU - Szemenyei, Márton AU - Szántó, Mátyás ED - Szirmay-Kalos, L ED - Renner, Gábor TI - DL-based Occlusion Detection in a Differentiable Simulation Environment T2 - X. Magyar Számítógépes Grafika és Geometria Konferencia PB - Neumann János Számítógép-tudományi Társaság CY - Budapest SN - 9789634218715 PY - 2022 SP - 143 EP - 150 PG - 8 UR - https://m2.mtmt.hu/api/publication/32872327 ID - 32872327 LA - English DB - MTMT ER - TY - CONF AU - Zoltán, Lőrincz AU - Szemenyei, Márton AU - Moni, Róbert TI - Imitation Learning for Generalizable Self-driving Policy with Sim-to-real Transfer T2 - ICLR 2022 Workshop on Generalizable Policy Learning in Physical World Proceedings PY - 2022 SP - 1 EP - 10 PG - 10 UR - https://m2.mtmt.hu/api/publication/32843087 ID - 32843087 LA - English DB - MTMT ER - TY - JOUR AU - Szemenyei, Márton AU - Reizinger, Patrik TI - Handling Realistic Noise in Multi-Agent Systems with Self-Supervised Learning and Curiosity JF - JOURNAL OF ARTIFICIAL INTELLIGENCE AND SOFT COMPUTING RESEARCH J2 - J ARTIF INTELL SOFT COMPUT RES VL - 12 PY - 2022 IS - 2 SP - 135 EP - 148 PG - 14 SN - 2083-2567 DO - 10.2478/jaiscr-2022-0009 UR - https://m2.mtmt.hu/api/publication/32754733 ID - 32754733 N1 - Funding Agency and Grant Number: NRDI Fund (TKP2020 NC) under Ministry for Innovation and TechnologyNational Research, Development & Innovation Office (NRDIO) - Hungary [BME-NC] Funding text: The research reported in this paper and carried out at BME has been supported by the NRDI Fund (TKP2020 NC, Grant No. BME-NC) based on the charter of bolster issued by the NRDI Office under the auspices of the Ministry for Innovation and Technology. AB - Most reinforcement learning benchmarks – especially in multi-agent tasks – do not go beyond observations with simple noise; nonetheless, real scenarios induce more elaborate vision pipeline failures: false sightings, misclassifications or occlusion. In this work, we propose a lightweight, 2D environment for robot soccer and autonomous driving that can emulate the above discrepancies. Besides establishing a benchmark for accessible multi-agent reinforcement learning research, our work addresses the challenges the simulator imposes. For handling realistic noise, we use self-supervised learning to enhance scene reconstruction and extend curiosity-driven learning to model longer horizons. Our extensive experiments show that the proposed methods achieve state-of-the-art performance, compared against actor-critic methods, ICM, and PPO. LA - English DB - MTMT ER - TY - CHAP AU - Szemenyei, Márton AU - Szántó, Mátyás ED - Kiss, Bálint ED - Szirmay-Kalos, László TI - Simulated Environment for Object Detection under Occlusion via Differential Rendering T2 - Proceedings of the Workshop on the Advances in Information Technology 2022 PB - OSZK CY - Budapest SN - 9789634218715 PY - 2022 SP - 39 EP - 45 PG - 7 UR - https://m2.mtmt.hu/api/publication/32670319 ID - 32670319 LA - English DB - MTMT ER -