Project Description

The Institute of High-Frequency Engineering and Electronics (IHE) at Karlsruhe Institute of Technology (KIT) is offering a Student Research Assistant (HiWi) position within an ongoing research project on neural network–driven CFAR (Constant False Alarm Rate) algorithms for automotive radar target detection.

The focus of the project is on complex highway scenarios, such as guardrails, roadside infrastructure, and dense static environments, where classical CFAR algorithms show limitations. The goal is to investigate and develop data-driven CFAR approaches that improve robustness and detection continuity under non-homogeneous clutter conditions.

Tasks and Responsibilities

Analysis of automotive radar data (e.g., Range-Doppler and Range-Angle representations) in complex highway environments

Manual and semi-automatic annotation of radar data for supervised and weakly supervised learning

Design and implementation of neural network–based CFAR and detection algorithms

Training, validation, and evaluation of neural networks for radar target detection

Comparison of classical CFAR methods with learning-based approaches

Documentation of results and support of scientific publications

Your Profile

background in signal processing and/or radar systems

Experience in neural networks and deep learning

Knowledge of computer vision / computational imaging concepts (e.g., CNNs, image-based detection, segmentation)

Programming skills in Python (PyTorch or TensorFlow preferred); MATLAB experience is a plus

Good command of English; German is not required

Institute of High-Frequency Engineering and Electronics (IHE)

as soon as possible

Jiayi Chen