Matija Mavsar
Papers
6
Total Citations
52
H-Index
5
About
Matija Mavsar is a leading researcher at the intersection of robotics and artificial intelligence, specializing in human-robot collaboration, intention recognition, and adaptive robotic systems. His work focuses on enabling robots to understand and predict human actions in real-time, making collaborative tasks safer and more efficient. Mavsar’s most impactful contribution is his development of recurrent neural network (RNN) architectures for intention recognition during dynamic human-robot handovers, as detailed in his highly cited 2021 paper (15 citations). He further advanced this field with simulation-aided handover prediction using recurrent image-to-motion networks (12 citations) and the vision-based RoverNet system for adaptive handovers (8 citations). His research also addresses practical industrial applications, including the integration of reconfigurable robotic workcells for automotive assembly (8 citations) and electronic waste recycling (2 citations). Notably, Mavsar has pioneered semi-supervised training methods using generative adversarial networks (GANs) to overcome data scarcity in cooperative robotics (7 citations). With a growing citation record and a focus on bridging theoretical deep learning with real-world robotic deployment, Mavsar is shaping the future of flexible, human-aware automation.
Research Focus
Key Achievements
Top Papers
- 1
- 2
- 3RoverNet: Vision-Based Adaptive Human-to-Robot Object Handovers8 citations · 2022
- 4
- 5
- 6