Alexandru Moraru

Tasks assigned: - design a module to improve the performance of angle estimation for close targets (<1m) - adapt misalignment estimation module for short-track calibration scenario - integrate the digital signal processing module into the complete software for cabin sensing radar project (adapt module to be compatible with complete software build chain, plan the delivery steps, align with the other stakeholders).

I applied machine learning to radar data in order to classify the road users (pedestrians, bicyclists and cars): - design and extract features from detection data - visualize and analyze these features and their dependency on position of the target - train machine learning models and evaluate their performance on a separate (hold out) test dataset - good classification accuracy

I was assigned with various tasks including: - developing an algorithm for trailer detection - developing a tracking algorithm for EMC testing - resolving customer requests for radar features - performing module test for radar features

I simulated in Matlab the Digital Signal Processing of a radar system that uses fast chirps including: - FFT for range and Doppler - noise modelling and detection (Neyman-Pearson criterion) using fixed and adaptive threshold (CFAR) - angle measurement (phase monopulse and digital beam forming)

Computer Vision Biometrics Machine Learning Deep Learning

Electrical Engineering