Evaluation of Different Generative Models to Support the Validation of Advanced Driver Assistance Systems
Validating the safety and reliability of automated driving systems is a critical challenge in the development of autonomous driving technology. Such systems must reliably replicate human driving behavior across scenarios of varying complexity and criticality. Ensuring this level of accuracy necessit...
Saved in:
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2025-05-01
|
| Series: | Applied Mechanics |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2673-3161/6/2/39 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1850156342800547840 |
|---|---|
| author | Manasa Mariam Mammen Zafer Kayatas Dieter Bestle |
| author_facet | Manasa Mariam Mammen Zafer Kayatas Dieter Bestle |
| author_sort | Manasa Mariam Mammen |
| collection | DOAJ |
| description | Validating the safety and reliability of automated driving systems is a critical challenge in the development of autonomous driving technology. Such systems must reliably replicate human driving behavior across scenarios of varying complexity and criticality. Ensuring this level of accuracy necessitates robust testing methodologies that can systematically assess performance under various driving conditions. Scenario-based testing addresses this challenge by recreating safety-critical situations at varying levels of abstraction, from simulations to real-world field tests. However, conventional parameterized models for scenario generation are often resource intensive, prone to bias from simplifications, and limited in capturing realistic vehicle trajectories. To overcome these limitations, the paper explores AI-based methods for scenario generation, with a focus on the cut-in maneuver. Four different approaches are trained and compared: Variational Autoencoder enhanced with a convolutional neural network (VAE), a basic Generative Adversarial Network (GAN), Wasserstein GAN (WGAN), and Time-Series GAN (TimeGAN). Their performance is assessed with respect to their ability to generate realistic and diverse trajectories for the cut-in scenario using qualitative analysis, quantitative metrics, and statistical analysis. Among the investigated approaches, VAE demonstrates superior performance, effectively generating realistic and diverse scenarios while maintaining computational efficiency. |
| format | Article |
| id | doaj-art-cfc9db11015b47d2a106129ee419e9be |
| institution | OA Journals |
| issn | 2673-3161 |
| language | English |
| publishDate | 2025-05-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Applied Mechanics |
| spelling | doaj-art-cfc9db11015b47d2a106129ee419e9be2025-08-20T02:24:34ZengMDPI AGApplied Mechanics2673-31612025-05-01623910.3390/applmech6020039Evaluation of Different Generative Models to Support the Validation of Advanced Driver Assistance SystemsManasa Mariam Mammen0Zafer Kayatas1Dieter Bestle2Mercedes-Benz AG, Kolumbusstr. 19+21, 71063 Sindelfingen, GermanyMercedes-Benz AG, Kolumbusstr. 19+21, 71063 Sindelfingen, GermanyDepartment of Engineering Mechanics and Vehicle Dynamics, Brandenburg University of Technology Cottbus-Senftenberg, Siemens-Halske-Ring 14, 03046 Cottbus, GermanyValidating the safety and reliability of automated driving systems is a critical challenge in the development of autonomous driving technology. Such systems must reliably replicate human driving behavior across scenarios of varying complexity and criticality. Ensuring this level of accuracy necessitates robust testing methodologies that can systematically assess performance under various driving conditions. Scenario-based testing addresses this challenge by recreating safety-critical situations at varying levels of abstraction, from simulations to real-world field tests. However, conventional parameterized models for scenario generation are often resource intensive, prone to bias from simplifications, and limited in capturing realistic vehicle trajectories. To overcome these limitations, the paper explores AI-based methods for scenario generation, with a focus on the cut-in maneuver. Four different approaches are trained and compared: Variational Autoencoder enhanced with a convolutional neural network (VAE), a basic Generative Adversarial Network (GAN), Wasserstein GAN (WGAN), and Time-Series GAN (TimeGAN). Their performance is assessed with respect to their ability to generate realistic and diverse trajectories for the cut-in scenario using qualitative analysis, quantitative metrics, and statistical analysis. Among the investigated approaches, VAE demonstrates superior performance, effectively generating realistic and diverse scenarios while maintaining computational efficiency.https://www.mdpi.com/2673-3161/6/2/39advanced driver assistance systemreal cut-in maneuversynthetic datavariational autoencodergenerative adversarial networksmachine learning |
| spellingShingle | Manasa Mariam Mammen Zafer Kayatas Dieter Bestle Evaluation of Different Generative Models to Support the Validation of Advanced Driver Assistance Systems Applied Mechanics advanced driver assistance system real cut-in maneuver synthetic data variational autoencoder generative adversarial networks machine learning |
| title | Evaluation of Different Generative Models to Support the Validation of Advanced Driver Assistance Systems |
| title_full | Evaluation of Different Generative Models to Support the Validation of Advanced Driver Assistance Systems |
| title_fullStr | Evaluation of Different Generative Models to Support the Validation of Advanced Driver Assistance Systems |
| title_full_unstemmed | Evaluation of Different Generative Models to Support the Validation of Advanced Driver Assistance Systems |
| title_short | Evaluation of Different Generative Models to Support the Validation of Advanced Driver Assistance Systems |
| title_sort | evaluation of different generative models to support the validation of advanced driver assistance systems |
| topic | advanced driver assistance system real cut-in maneuver synthetic data variational autoencoder generative adversarial networks machine learning |
| url | https://www.mdpi.com/2673-3161/6/2/39 |
| work_keys_str_mv | AT manasamariammammen evaluationofdifferentgenerativemodelstosupportthevalidationofadvanceddriverassistancesystems AT zaferkayatas evaluationofdifferentgenerativemodelstosupportthevalidationofadvanceddriverassistancesystems AT dieterbestle evaluationofdifferentgenerativemodelstosupportthevalidationofadvanceddriverassistancesystems |