Vladimiros Sterzentsenko

n the current role, the requirements of a growing startup are being met by supporting multiple features rapidly. This position involves technical leadership within the company and participation in nearly every aspect of the real-time Motion Capture solution offered. Responsibilities and finished tasks include:

• Re-designed and refactored prototype Motion Capture software to enable scalability, security and performance
• Designing and implementing end-to-end multi-camera calibration solution
• Implementing multiple sensor back-ends to work seamlessly with current software solution
• Implementing high throughput computer vision algorithms
• Designing and developing software UI and UX

At Soar, the quality of the real-time streamed 3D mesh was significantly improved and the software's performance was enhanced through the use of high-throughput image processing algorithms and advanced rendering techniques. Key contributions included:

• Enhancing multi-camera calibration results
• Designing and implementing high-throughput solutions to improve sensor data
• Playing major role in GPU-accelerated refactoring company's software
• Applying advanced rendering techniques to enhance visual quality of streamed mesh
• Implementing offline processing pipeline

At CERTH-ITI, I served as a Computer Vision Engineer and Research Scientist, actively contributing to various projects and authoring several publications. Key responsibilities included::

• Redesigning, enhancing, and implementing lab's multi-camera performance capture system
• Designing and implementing multi-camera calibration systems for depth and color sensors
• Developing real-time deep-learning solutions for depth sensor data filtering
• Creating generative deep-learning methods for panoramic image in-painting
• Working on deep-learning methods for indoor scene understanding