Constantin Patsch

Papers

Total Citations

H-Index

About

Constantin Patsch is a researcher at the forefront of advancing human-robot collaboration through cutting-edge work in temporal action segmentation and contrastive learning. His primary research areas include computer vision, human activity understanding, and non-verbal communication in robotics. Patsch’s major contribution lies in developing novel machine learning frameworks that enable robots to interpret complex sequences of human actions, a critical capability for seamless, intuitive human-robot interaction. His most-cited paper, "TSCL: Timestamp Supervised Contrastive Learning for Action Segmentation" (2024), introduces a pioneering approach that leverages timestamp supervision to improve the accuracy of action segmentation, allowing robots to recognize long-term dependencies and underlying human intentions. Although early in his career, this work has already garnered 2 citations, signaling growing impact in the field. Patsch’s research is particularly notable for its practical implications in robotic assistance and collaborative tasks, where understanding non-verbal cues is essential. His achievements underscore a commitment to bridging the gap between human behavior and autonomous systems, making him a promising voice in the evolution of intelligent, context-aware robotics.