A5004645317- Advanced Optical Sensing Technologies
- Advanced SAR Imaging Techniques
- Autonomous Vehicle Technology and Safety
- Geophysical Methods and Applications
- Underwater Acoustics Research
- Advanced Neural Network Applications
- Transportation Safety and Impact Analysis
- Indoor and Outdoor Localization Technologies
- Target Tracking and Data Fusion in Sensor Networks
- Radar Systems and Signal Processing
- Anomaly Detection Techniques and Applications
- GNSS positioning and interference
- Advanced Clustering Algorithms Research
- Remote Sensing and LiDAR Applications
- Fault Detection and Control Systems
- Random lasers and scattering media
- Sparse and Compressive Sensing Techniques
- Traffic Prediction and Management Techniques
- Microwave Imaging and Scattering Analysis
- Ocular and Laser Science Research
- Bayesian Methods and Mixture Models
- Engineering Applied Research
- Risk and Safety Analysis
- Automotive and Human Injury Biomechanics
- IoT and GPS-based Vehicle Safety Systems
Mercedes-Benz (Germany)
2020-2022
Daimler (Germany)
2018-2019
University of Kassel
2019
A new automotive radar data set with measurements and point-wise annotations from more than four hours of driving is presented. Data provided by series sensors mounted on one test vehicle were recorded the individual detections dynamic objects manually grouped to clusters labeled afterwards. The purpose this enable development novel (machine learning-based) perception algorithms focus moving road users. Images sequences captured using a documentary camera. For evaluation future object...
Abstract Automotive radar perception is an integral part of automated driving systems. Radar sensors benefit from their excellent robustness against adverse weather conditions such as snow, fog, or heavy rain. Despite the fact that machine-learning-based object detection traditionally a camera-based domain, vast progress has been made for lidar sensors, and also catching up. Recently, several new techniques using machine learning algorithms towards correct classification moving road users in...
Radar sensors are an important part of driver assistance systems and intelligent vehicles due to their robustness against all kinds adverse conditions, e.g., fog, snow, rain, or even direct sunlight. This is achieved by a substantially larger wavelength compared light-based such as cameras lidars. As side effect, many surfaces act like mirrors at this wavelength, resulting in unwanted ghost detections. In article, we present novel approach detect these objects applying data-driven machine...
Conventional sensor systems record information about directly visible objects, whereas occluded scene components are considered lost in the measurement process. Non-line-of-sight (NLOS) methods try to recover such hidden objects from their indirect reflections - faint signal components, traditionally treated as noise. Existing NLOS approaches struggle these low-signal outside lab, and do not scale large-scale outdoor scenes high-speed motion, typical automotive scenarios. In particular,...
Radar-based road user detection is an important topic in the context of autonomous driving applications. The resolution conventional automotive radar sensors results a sparse data representation which tough to refine during subsequent signal processing. On other hand, new sensor generation waiting wings for its application this challenging field. In article, two different generations are evaluated against each other. evaluation criterion performance on moving object and classification tasks....
The classification of individual traffic participants is a complex task, especially for challenging scenarios with multiple road users or under bad weather conditions. Radar sensors provide an - respect to well established camera systems orthogonal way measuring such scenes. In order gain accurate results, 50 different features are extracted from the measurement data and tested on their performance. From these suitable subset chosen passed random forest long short-term memory (LSTM)...
Radar-based road user classification is an important yet still challenging task towards autonomous driving applications. The resolution of conventional automotive radar sensors results in a sparse data representation which tough to recover by subsequent signal processing. In this article, classifier ensembles originating from one-vs-one binarization paradigm are enriched one-vs-all correction classifiers. They utilized efficiently classify individual traffic participants and also identify...
Annotating automotive radar data is a difficult task. This article presents an automated way of acquiring labels which uses highly accurate and portable global navigation satellite system (GNSS). The proposed discussed besides revision other label acquisitions techniques problem description manual annotation. concludes with systematic comparison conventional hand labeling automatic acquisition. results show clear advantages the method without relevant loss in accuracy. Minor changes can be...
Radar sensors provide a unique method for executing environmental perception tasks towards autonomous driving. Especially their capability to perform well in adverse weather conditions often makes them superior other such as cameras or lidar. Nevertheless, the high sparsity and low dimensionality of commonly used detection data level is major challenge subsequent signal processing. Therefore, points are merged order form larger entities from which more information can be gathered. The...
Radar sensors have a long tradition in advanced driver assistance systems (ADAS) and also play major role current concepts for autonomous vehicles. Their importance is reasoned by their high robustness against meteorological effects, such as rain, snow, or fog, the radar's ability to measure relative radial velocity differences via Doppler effect. The cause these advantages, namely large wavelength, one of drawbacks radar sensors. Compared camera lidar sensor, lot more surfaces typical...
Baseline generation for tracking applications is a difficult task when working with real world radar data. Data sparsity usually only allows an indirect way of estimating the original tracks as most objects' centers are not represented in This article proposes automated acquiring reference trajectories by using highly accurate hand-held global navigation satellite system (GNSS). An embedded inertial measurement unit (IMU) used orientation and motion behavior. contains two major...
A new automotive radar data set with measurements and point-wise annotations from more than four hours of driving is presented. Data provided by series sensors mounted on one test vehicle were recorded the individual detections dynamic objects manually grouped to clusters labeled afterwards. The purpose this enable development novel (machine learning-based) perception algorithms focus moving road users. Images sequences captured using a documentary camera. For evaluation future object...