Dr.-Ing. Maryam Asadzadeh

• Led a team delivering AI solutions for smart manufacturing • Built deep learning models for object recognition & time series analysis for quality control • Optimized edge/mobile inference for faster, more accurate results

• Built AI-driven computer vision solutions for quality control in automotive manufacturing, such as real-time defect detection deployed models on production lines. • Optimized production line inspections using machine vision systems. • Led company’s AI Cup team in an international deep learning competition.

Design and develop numerous computer vision and training AI projects such as: • Human pose estimation inside the vehicle cabin from top view • Object detection/ segmentation within large UAV images • Face, hand and object recognition and tracking in video retail management • Novel automatic annotation, imaging segmentation based on one-shot object detection • Managing a team of computer vision engineers and mentoring interns

• Implement novel computer vision and deep learning algorithms for CCTV video understanding. • Detect any human and vehicle activities to reduce false alarms in complex and low quality CCTV videos • Research collaboration with external researchers on camera shakiness detection, image enhancement, super resolution and blurriness leading to publish scientific articles • Supervision of master thesis with title ‘object detection for CCTV monitoring in construction environments’ at university of applied sciences and arts Dortmund • Managing AI team of computer vision engineer and scrum master

• Conducting state of the art research and building innovated computer vision algorithms via rapid prototyping for autonomous vehicle and micro mobility applications such as: Parking zone, lane, scooter, manhole, pedestrian, helmet and multi rider detection • Working with hardwares and optimisation approaches to perform real-time predictions on edge devices such as raspberry pi and jetson nano

• Design and implement various computer vision and machine learning projects such as: • Scene classification and recognition in surveillance applications such as flood detection • Image size reduction techniques for internet application based on saliency • Zone segmentation in autonomous drone safe landing • License plate identification based on CNN - RNN • Face and Family classification based on geometric and entropy models • Handwriting recognition for age and gender identification, dataset creation • Published ten scientific papers in top computer vision journals and conferences • Supervision of Master thesis students

• Modules taught: computer vision, advance network, network security, computer programming python and C++, security of information, operation system, discrete structures to name a few

My core area of research: image processing and applied machine learning and deep learning ‘Twelve’ published and ‘One’ under review papers during 3-years PhD in top journals and conferences Thesis title: Image Classification and Segmentation for Efficient Surveillance Applications Project subjects: licence plate recognition, smart UAV safe landing, scene classification, face recognition, handwriting analysis, image compression to name a few

Thesis title: SSL/TLS Security Protocol Improvement Based On Trust & Reputation Model