Pavol Jancura

About

Pavol Jancura is a researcher advancing the frontiers of robot learning and human-robot interaction, with a focus on trajectory prediction and inverse reinforcement learning (IRL). His work addresses critical challenges in enabling mobile robots to navigate safely and efficiently in dynamic, multi-environment settings. In his highly cited 2022 paper, "Continual Pedestrian Trajectory Learning With Social Generative Replay" (20 citations), Jancura tackles the problem of pedestrian motion pattern variability across environments. He introduces a continual learning framework that uses social generative replay to prevent catastrophic forgetting, allowing robots to adapt to new spaces without losing prior knowledge—a key step toward robust, lifelong deployment. His 2022 work on "Model-free inverse reinforcement learning with multi-intention, unlabeled, and overlapping demonstrations" (2 citations) further expands the IRL toolkit, solving a complex problem where demonstrations come from multiple unknown experts with shared behaviors. This model-free approach eliminates the need for intention labels, making it practical for real-world scenarios like autonomous driving or assistive robotics. Jancura’s contributions are shaping how robots learn from human behavior, with implications for safer, more adaptive autonomous systems.

Research Focus

Key Achievements

Top Papers

1. 1
   Continual Pedestrian Trajectory Learning With Social Generative Replay

   20 citations · 2022
2. 2
   Model-free inverse reinforcement learning with multi-intention, unlabeled, and overlapping demonstrations

   2 citations · 2022

Key Collaborators

Contact & Links