SENSE, MAP AND ACT IN REAL TIME
Robots, drones, trains and cars need to see, map and decide in real time. We build the perception and mapping that lets them act safely on their own.
Topics
- Vision-language models for autonomous trains Mohammadjavad Ghorbanalivakili, Ashley Varghese
- Autonomous mapping agents for drones Looking for students
- 3D-grounded vision-language-action models Looking for students
Key papers
- Reasoning-guided Ego-path Segmentation for Autonomous Trains using Vision-Language Models Ghorbanalivakili, M., Varghese, A., Sohn, G. · ISPRS Archives, 2026
- The Emerging Role of Vision-Language Models in the Automation of Railway Asset Management: A Review and Future Perspective Varghese, A., Ghorbanalivakili, M., Sohn, G. · ISPRS Archives, XLIX-B2-2026, 745–753, 2026
- Polarization-Aware Segmentation for Camouflaged Threat Detection from UAVs Korny, Y., Yoo, S., Sohn, G. · ISPRS Archives, XLIX-B2-2026, 705–713, 2026
