Building a Training Dataset for Machine Learning, Radar-Based Pedestrian Detection

Zoltán Rózsás; István Lakatos; Tamás Péter

doi:10.30939/ijastech..1756258

Building a Training Dataset for Machine Learning, Radar-Based Pedestrian Detection

Abstract

Abstract: Radar-based pedestrian detection is a key enabler of advanced driver-assistance systems (ADAS) and future autonomous driving functions, particularly under conditions where vision sensors are limited by poor lighting or adverse weather. In this study, we present a structured dataset collected at the ZalaZONE proving ground, a state-of-the-art automotive testing facility in Hungary, using a Continental ARS 408‑21 automotive radar operating at 77 GHz . The dataset contains static radar measurements of both a real human subject and a pedestrian dummy, recorded at multiple controlled distances (5 m and 10 m) and orientations (front-facing and side-facing). Each radar scan includes radar cross-section (RCS), object distance, relative velocity, and metadata retrieved directly from the vehicle’s CAN bus interface. The results demonstrate that RCS values strongly depend on target type, orientation, and distance. Real pedestrians produce significantly higher variance due to posture, clothing materials, and micro‑movements, whereas dummy mannequins exhibit stable, narrow RCS distributions. Interestingly, measurements at longer distances show less negative RCS values, likely influenced by multipath reflections and environmental dispersion effects. The dataset is fully labeled and formatted for direct use in supervised machine learning pipelines, supporting classification models such as logistic regression, support vector machines, and neural networks. It also pro-vides a foundation for future extensions with dynamic scenes and temporal sequence modeling, enabling the development of more robust and generalizable radar-based pedestrian detection algorithms. By combining controlled measurements with realistic environmental variability, this dataset contributes to the advancement of radar sensing technologies for safe and reliable autonomous driving. Quantitatively, the mean RCS difference between real pedestrians and dummy targets reached 3.8 dB at 5 m and 4.5 dB at 10 m, confirming the discriminative potential of the dataset for classification tasks.

Keywords

Supporting Institution

Zala Zona and Szechenyi Istvan University Győr

References

[1] Harlow K, Jang H, Barfoot TD, Kim A, Heckman C. A new wave in robotics: survey on recent mmWave radar applications in robot-ics. IEEE Trans Robot. 2024;40:4544-4560. https://doi.org/10.1109/TRO.2024.3463504
[2] Wei Z, Zhang F, Chang S, Liu Y, Wu H, Feng Z. mmWave radar and vision fusion for object detection in autonomous driving: a re-view. Sensors (Basel). 2022;22(7):2542. https://doi.org/10.3390/s22072542
[3] Han Z, Wang J, Xu Z, Yang S, He L, Xu S, et al. 4D millimeter-wave radar in autonomous driving: a survey. arXiv [Preprint]. 2023 Jun 7. Available from: https://arxiv.org/abs/2306.04242
[4] Skolnik MI. Radar handbook. 3rd ed. New York: McGraw-Hill Education; 2008.
[5] Manzoni M, Tebaldini S, Monti-Guarnieri AV, Prati CM. Multi-path in automotive MIMO SAR imaging. IEEE Trans Geosci Remote Sens. 2023;61:1-12. https://doi.org/10.1109/TGRS.2023.3240705
[6] Li Y, Feng Z. Human motion recognition using millimeter-wave radar. IEEE Geosci Remote Sens Lett. 2018;15(12):1917-1921. https://doi.org/10.1109/LGRS.2018.2859365
[7] Roh Y, Park S. Micro-Doppler-based human motion classification using FMCW radar. IEEE Sens J. 2020;20(9):4905-4917. https://doi.org/10.1109/JSEN.2020.2969863
[8] Palffy A, Kooij JFP, Gavrila DM. Detecting darting out pedestri-ans with occlusion aware sensor fusion of radar and stereo camera. IEEE Trans Intell Veh. 2023;8(2):1459-1472. https://doi.org/10.1109/TIV.2022.3220435

Details

Primary Language

English

Subjects

Automotive Mechatronics and Autonomous Systems

Journal Section

Research Article

Authors

Zoltán Rózsás ^*
0000-0002-9484-4313
Hungary

István Lakatos
0000-0002-3855-7379
Hungary

Tamás Péter
0000-0001-5386-5188
Hungary

Early Pub Date

December 1, 2025

Publication Date

December 17, 2025

Submission Date

August 10, 2025

Acceptance Date

November 19, 2025

Published in Issue

Year 2025 Volume: 9 Number: 1st Future of Vehicles Conf.

DOI

https://doi.org/10.30939/ijastech..1756258

IZ

https://izlik.org/JA36AD75JW

Cite

RIS / Bibtex

APA

Rózsás, Z., Lakatos, I., & Péter, T. (2025). Building a Training Dataset for Machine Learning, Radar-Based Pedestrian Detection. International Journal of Automotive Science And Technology, 9(1st Future of Vehicles Conf.), 72-76. https://doi.org/10.30939/ijastech..1756258

AMA

1.Rózsás Z, Lakatos I, Péter T. Building a Training Dataset for Machine Learning, Radar-Based Pedestrian Detection. IJASTECH. 2025;9(1st Future of Vehicles Conf.):72-76. doi:10.30939/ijastech.1756258

Chicago

Rózsás, Zoltán, István Lakatos, and Tamás Péter. 2025. “Building a Training Dataset for Machine Learning, Radar-Based Pedestrian Detection”. International Journal of Automotive Science And Technology 9 (1st Future of Vehicles Conf.): 72-76. https://doi.org/10.30939/ijastech. 1756258.

EndNote

Rózsás Z, Lakatos I, Péter T (December 1, 2025) Building a Training Dataset for Machine Learning, Radar-Based Pedestrian Detection. International Journal of Automotive Science And Technology 9 1st Future of Vehicles Conf. 72–76.

IEEE

[1]Z. Rózsás, I. Lakatos, and T. Péter, “Building a Training Dataset for Machine Learning, Radar-Based Pedestrian Detection”, IJASTECH, vol. 9, no. 1st Future of Vehicles Conf., pp. 72–76, Dec. 2025, doi: 10.30939/ijastech..1756258.

ISNAD

Rózsás, Zoltán - Lakatos, István - Péter, Tamás. “Building a Training Dataset for Machine Learning, Radar-Based Pedestrian Detection”. International Journal of Automotive Science And Technology 9/1st Future of Vehicles Conf. (December 1, 2025): 72-76. https://doi.org/10.30939/ijastech. 1756258.

JAMA

1.Rózsás Z, Lakatos I, Péter T. Building a Training Dataset for Machine Learning, Radar-Based Pedestrian Detection. IJASTECH. 2025;9:72–76.

MLA

Rózsás, Zoltán, et al. “Building a Training Dataset for Machine Learning, Radar-Based Pedestrian Detection”. International Journal of Automotive Science And Technology, vol. 9, no. 1st Future of Vehicles Conf., Dec. 2025, pp. 72-76, doi:10.30939/ijastech. 1756258.

Vancouver

1.Zoltán Rózsás, István Lakatos, Tamás Péter. Building a Training Dataset for Machine Learning, Radar-Based Pedestrian Detection. IJASTECH. 2025 Dec. 1;9(1st Future of Vehicles Conf.):72-6. doi:10.30939/ijastech. 1756258

Cited By

Analysis of Following Distance Measurement and Compliance in Road Traffic

Engineering Perspective

https://doi.org/10.64808/engineeringperspective.1791630