Csaba Beleznai
Papers
5
Total Citations
16
H-Index
2
About
Csaba Beleznai is a leading researcher at the intersection of computer vision, robotics, and human-machine collaboration. His work focuses on enabling autonomous systems to perceive and interact with complex, unstructured environments—particularly in industrial and safety-critical settings. A central contribution is his pioneering integration of large vision-language models into robotic control, exemplified by "TalkWithMachines" (2024, 7 citations), which allows robots to understand natural language commands and translate their internal state into intuitive human feedback. Beleznai has also made significant strides in automated logistics, developing robust methods for pallet detection and 3D pose estimation from synthetic data (2025, 4 citations; 2023, 2 citations), overcoming occlusion challenges that plague real-world deployment. His earlier work on real-time human detection using contour template matching (2011, 2 citations) laid groundwork for visual surveillance applications. Notably, his "PrimitivePose" framework (2023) tackles the fundamental challenge of predicting 3D bounding boxes for unseen objects without requiring pre-existing CAD models. With a career spanning foundational perception algorithms to cutting-edge LLM-robot interfaces, Beleznai’s research directly addresses the practical hurdles of automating diverse production and transport workflows.
Research Focus
Key Achievements
Top Papers
- 1
- 2
- 3
- 4
- 5