Qiao Qiao

Title: Uncalibrated camera pose estimation of monocular RGB image of a single person. Goal: Warp image with estimated camera parameters to a desired perspective that would improve 3D model fitting. Design the whole pipeline, experiment with different approaches, research on 2D/3D human pose estimation, build up customized datasets, render images from 3D models. Design, train, evaluate and test deep neural networks on synthetic and real images. (Python, PyTorch)

3D scanning, 3D human model processing, body reconstruction, and measurement extraction with Artec Studio. Optimize the whole data collection process such as writing Python scripts to automatically read measurements from XML files, reducing time from 10min to 10s. Research on virtual try-on, size recommendation. Develop video obfuscation microservice (Human segmentation, Docker, Flask, PostgreSQL, Google Cloud Storage) Familiar with start-up environment, agile project management, hybrid/remote work style.

Tutorin der Vorlesung „Grundlagen der künstlichen Intelligenz“ Python Programmierung für ein Route Planner in commonroad

- Python Programmieren - Produktprüfung - Herstellung elektromechanischer Geräte

Automatisch Übertragung von Safe-Roboter- Daten in die KUKA-Safe-Robot Dokumentation mit VBA Programmierung

- Robotersimulation im Maschinen- und Anlagenbau - Einarbeitung in das Simulationssystem Process Simulate von Siemens - Roboterkinematisierung und Programmierung ( spez. LBR iiwa)

Computer vision, Machine Learning, AI, Deep Learning

Informatik, Control Theory, Elektrotechnik