Valsamis Ntouskos
Papers
7
Total Citations
98
H-Index
6
About
Valsamis Ntouskos is a leading researcher at the intersection of robotics, computer vision, and human-robot interaction. His work centers on enabling robots to perceive, understand, and anticipate human actions in real-world environments, with a particular focus on assistive robotics and disaster response. Ntouskos’s most impactful contribution is his involvement in the TRADR project, where he helped deploy ground and aerial robots in the aftermath of the 2016 Amatrice earthquake in Italy. This pioneering work, cited 55 times, demonstrated how robots can collect critical data to create 3D textured models of damaged heritage sites, showcasing the practical value of field robotics in crisis scenarios. Beyond disaster response, Ntouskos has made significant advances in human motion analysis. He introduced the concept of “motion flux” to automatically discover and recognize motion primitives from 3D pose data, a framework that enables robots to break down complex human activities into learnable components. His subsequent research on next-action prediction and goal anticipation has been instrumental in developing robots that can proactively assist humans by inferring their intentions. With a series of 2019 papers on execution monitoring and visual search for assistive tasks, Ntouskos continues to push the boundaries of how robots can seamlessly collaborate with people in dynamic settings.
Research Focus
Key Achievements
Top Papers
- 1
- 2Discovery and recognition of motion primitives in human activities15 citations · 2019
- 3Help by Predicting What to Do6 citations · 2019
- 4Anticipating Next Goal for Robot Plan Prediction6 citations · 2019
- 5Deep Execution Monitor for Robot Assistive Tasks6 citations · 2019
- 6Visual search and recognition for robot task execution and monitoring6 citations · 2019
- 7Human motion primitive discovery and recognition.4 citations · 2017