New Challenges for Deep Neural Networks in Automotive Radar Perception

Scheiner, Nicolas; Weishaupt, Fabio; Tilly, Julius F.; Dickmann, Jürgen

doi:10.1007/978-3-658-34752-9_14

Cited by 5 publications

(5 citation statements)

References 37 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Dealing with sensor limitations and uncertainties: Sensors used in perception systems have inherent limitations and uncertainties [138,139]. Cameras may have a limited eld of view or be affected by lighting conditions [140], lidar sensors may face challenges in detecting certain materials or have a limited range [141], and radar sensors may have di culty distinguishing between objects nearby [142].…”

Section: Discussionmentioning

confidence: 99%

Perception System Architecture for Self-Driving Vehicles: A Cyber- Physical Systems Framework

Babaei,

Riahinia,

Ebadati

et al. 2023

Preprint

View full text Add to dashboard Cite

This research is seeking for design the autonomous vehicle's perception system from the perspective of cyber-physical system. The perception system is considered a cyber and physical component. This research takes a systemic approach. Unified Modeling Language (UML) has been applied to design perception systems. The perceptual system was created using the 3 common UML diagrams. The findings show that UML can be used as a powerful tool for modeling every type of complex system. The sensors utilized in the tasks of the perception layer are shown in use case diagrams. To recognize surrounding things and avoid obstacles, a variety of technologies including cameras, LiDAR, RADAR, ultrasonic sensors, GPS, and GNSS are utilized in 2/3-d object detection, speed and distance detection, localization, and mapping. The class diagram displays the perception system's structural elements. The perception system's task execution was shown in the sequence diagram. The engineers can improve the functionality and security of autonomous vehicles by using UML to create a precise and organized description of the perception system. The model in which designed in this paper can improve autonomous vehicles' perception system performance, driving safety, and self-confidence of vehicles. this model is an innovative model based on the cyber-physical systems approach and with the UML modeling method. other models in the literature are pipeline modular and/or see the perception layer tasks in isolation.

show abstract

Section: Discussionmentioning

confidence: 99%

Perception System Architecture for Self-Driving Vehicles: A Cyber- Physical Systems Framework

Babaei,

Riahinia,

Ebadati

et al. 2023

Preprint

View full text Add to dashboard Cite

show abstract

“…Scheiner et al in [9], paid particular attention to the quality of input data to the radar perception process and the use of deeper and more complex neural network structures. For sparsity, high-resolution processing or the use of low-level data layers and polarimetric radars are used.…”

Section: Introductionmentioning

confidence: 99%

Radar Perception of Multi-Object Collision Risk Neural Domains during Autonomous Driving

Lisowski

2024

Electronics

View full text Add to dashboard Cite

The analysis of the state of the literature in the field of methods of perception and control of the movement of autonomous vehicles shows the possibilities of improving them by using an artificial neural network to generate domains of prohibited maneuvers of passing objects, contributing to increasing the safety of autonomous driving in various real conditions of the surrounding environment. This article concerns radar perception, which involves receiving information about the movement of many autonomous objects, then identifying and assigning them a collision risk and preparing a maneuvering response. In the identification process, each object is assigned a domain generated by a previously trained neural network. The size of the domain is proportional to the risk of collisions and distance changes during autonomous driving. Then, an optimal trajectory is determined from among the possible safe paths, ensuring control in a minimum of time. The presented solution to the radar perception task was illustrated with a computer simulation of autonomous driving in a situation of passing many objects. The main achievements presented in this article are the synthesis of a radar perception algorithm mapping the neural domains of autonomous objects characterizing their collision risk and the assessment of the degree of radar perception on the example of multi-object autonomous driving simulation.

show abstract

“…The application of deep learning in radar perception has drawn extensive attention from autonomous driving researchers. In the past two years, a number of review papers [3][4][5][6][7][8] have been published in this field. Zhou et al [3] categorise radar perception tasks into dynamic target detection and static environment modelling.…”

Section: Introductionmentioning

confidence: 99%

“…They also introduce approaches for radar and camera fusion based on the classical taxonomy of the data-level, feature-level, and decision-level. Scheiner et al [5] discuss the information sparsity problem and labelling challenge in learning-based radar perception. Three strategies are recommended to increase radar data density, including the use of pre-CFAR data, the use of high-resolution radar sensors, and the use of polarisation information.…”

Section: Introductionmentioning

confidence: 99%

Towards Deep Radar Perception for Autonomous Driving: Datasets, Methods, and Challenges

Zhou

Liu

Zhao

et al. 2022

Sensors

View full text Add to dashboard Cite

With recent developments, the performance of automotive radar has improved significantly. The next generation of 4D radar can achieve imaging capability in the form of high-resolution point clouds. In this context, we believe that the era of deep learning for radar perception has arrived. However, studies on radar deep learning are spread across different tasks, and a holistic overview is lacking. This review paper attempts to provide a big picture of the deep radar perception stack, including signal processing, datasets, labelling, data augmentation, and downstream tasks such as depth and velocity estimation, object detection, and sensor fusion. For these tasks, we focus on explaining how the network structure is adapted to radar domain knowledge. In particular, we summarise three overlooked challenges in deep radar perception, including multi-path effects, uncertainty problems, and adverse weather effects, and present some attempts to solve them.

show abstract

New Challenges for Deep Neural Networks in Automotive Radar Perception

Cited by 5 publications

References 37 publications

Perception System Architecture for Self-Driving Vehicles: A Cyber- Physical Systems Framework

Perception System Architecture for Self-Driving Vehicles: A Cyber- Physical Systems Framework

Radar Perception of Multi-Object Collision Risk Neural Domains during Autonomous Driving

Towards Deep Radar Perception for Autonomous Driving: Datasets, Methods, and Challenges

Contact Info

Product

Resources

About