Georgios Sopidis
Procomcure Biotech (Austria), Johannes Kepler University of Linz
Papers
3
Total Citations
18
H-Index
3
About
Georgios Sopidis is a researcher at the forefront of applied deep learning and sensor-based monitoring, with key contributions spanning human activity recognition, industrial manufacturing, and critical infrastructure safety. His work uniquely bridges the gap between raw sensor data and actionable insights, particularly through the use of inertial measurement units (IMUs) and multi-sensor robotic systems. In his most cited work, "Counting Activities Using Weakly Labeled Raw Acceleration Data," Sopidis tackles the fundamental challenge of variable-duration activity detection, introducing a variable-length sequence approach with deep learning that eliminates the constraints of fixed window sizes—a breakthrough for real-world applications like gesture counting. This paper has garnered 7 citations, reflecting its immediate relevance. His impact extends to industrial process optimization, as seen in "Analyzing Arc Welding Techniques improves Skill Level Assessment," where he applies machine learning to automate quality assessment in manual metal arc welding, a task still reliant on human expertise. Notably, his 2024 paper on "First Measurement Campaign by a Multi-Sensor Robot for the Lifecycle Monitoring of Transformers" addresses a critical gap in post-event forensics for electrical grid assets, using robotic sensing to capture data from rare but destructive transformer failures. With a growing citation footprint and a focus on practical, high-stakes applications, Sopidis is establishing himself as a key innovator in sensor-driven AI for both human performance and industrial safety.
Research Focus
Key Achievements
Top Papers
- 1
- 2
- 3