Armin Razavi

• Built Python/C++ tooling and workflows for processing large X-ray/CT datasets used in AI model development, validation, benchmarking, and release decisions. • Developed and optimized machine-learning and deep-learning based object-detection workflows with TensorFlow/PyTorch and TensorRT in industrial AI systems. • Improved training-data generation, annotation QA, metadata handling, coverage analysis, and data-quality checks for model-development teams. • Automated recurring engineering and data-prepara

Developing advanced deep learning complete end-to-end algorithms for automatic threat detection in X-ray and CT baggage images. Provide guidance and support to team members in developing essential tools and scripts to automate data engineering and data processing pipelines, leveraging Docker-based DevOps processes. Play an active role in project planning and strategy development, while also assessing and selecting suitable technologies and tools to promote innovation and ensure successful project exec

Creating robot vision algorithms utilizing Robot Operating System for depth sensors (GARotics Project ) Implementing scene semantic recognition and segme ntation techniques using point cloud data. Engaged as a research associate and provided assistance as a professor's assistant.

Creating robot vision algorithms utilizing Robot Operating System for depth sensors (GARotics Project) Implementing scene semantic recognition and segmentation techniques using point cloud data. Engaged as a research associate and provided assistance as a professor's assistant.

Devel o ping machine learning - based computer - aided diagnosis (CAD) systems for breast lesion cla ssification specifically in dynamic contrast - enhanced MRI (DCE - MRI) images (ASSURE Project ) Completing the master's thesis

Developing machine learning-based computer-aided diagnosis (CAD) systems for breast lesion classification specifically in dynamic contrast-enhanced MRI (DCE-MRI) images (ASSURE Project) Completing the master's thesis

Developing a semi-automatic document anonymization, incorporating NLP, HCI, and data mining techniques, leveraging GATE (General Architecture for Text Engineering) toolkit (QualiService Project)

Developing a semi-automatic document anonymization, incorporating NLP, HCI, and data mining techniques, leveraging GATE (General Architecture for Text Engineering) toolkit (qualiservice.org)

Multimedia software development (Education and Awareness Department) Research Assistant http://www.nsec.ir/

Lecturing C++ programming, 3D Studio Max and Adobe Flash

GPS Vehicle Tracking systems http://www.irsystem.org/

R&D, developing multimedia educational software solutions (Education and Awareness Department)

Master’s Thesis: “Automatic Classification of Non-Mass Breast Lesions in DCE-MR Images” in cooperation with Fraunhofer MEVIS. Available online at http://cgvr.cs.uni-bremen.de/theses/finishedtheses/breast_lesions/index.shtml Master’s Project: “Natural User Interaction in Cars” (Project Tarpan) in c

Software system analysis and agile system design