Software Developer (Computer Vision / ML)

An aspiring Software Developer with a focus on Computer Vision, Object Detection, and Machine Learning for Advanced Driver-Assistance Systems (ADAS). He developed and optimized Deep Learning models (RTMDet, LSTM, Transformer, 1D-CNN) for real-time drone detection and vehicle trajectory prediction. His expertise encompasses data augmentation, feature extraction, and model optimization. He contributed to publications and a patent application in the field of ML-based ADAS.

Development, training, and optimization of an RTMDet-based object detection model for real-time drone detection with significant mAP increase; Systematic comparison of CNN architectures (YOLO, Faster R-CNN, RTMDet) for micro-drone detection; Application of feature extraction, data augmentation (including synthetic data generation with Unreal Engine), and model optimization; Generation of 220 synthetic driving scenarios with IPG CarMaker for ML-based trajectory prediction; Conception, implementation, and evaluation of LSTM, Transformer, and 1D-CNN architectures for vehicle trajectory prediction; Evaluation and optimization of model performance based on prediction accuracy, inference latency, and driving comfort; Publication of research results in ATZ worldwide and submission of a patent application.