Robot Learning
Imitation learning, reinforcement learning, model-based learning and generalist robot policies.
We build learning-enabled robotic systems that perceive, reason and act in complex real-world environments.
HURLAB investigates how robots can build useful representations of the world, make decisions under uncertainty and learn robust behaviors from data, models and interaction.
Imitation learning, reinforcement learning, model-based learning and generalist robot policies.
Decision-making, motion understanding and reliable autonomy in dynamic environments.
Multimodal sensing with cameras, RGB-D and LiDAR for semantic scene understanding.
Place recognition, topological mapping and long-term spatial representations.
Segmentation, detection, object pose estimation and 3D environment understanding.
Gesture understanding, collaborative robotics and learning grounded actions from people.
HURLAB brings together faculty collaborators and graduate researchers working across artificial intelligence, computer vision and robotics.
PhD Candidate
PhD Candidate
MSc Student
MSc Student
MSc Student
MSc Alumnus
MSc Alumna
Research on multimodal road understanding, semantic grid prediction, trajectory forecasting and perception for autonomous systems.
İsmail Emre Canıtez and Özgür Erkent
Aysu Aylin Kaplan and Özgür Erkent
Hüseyin Arslan and Özgür Erkent
Özgür Erkent and collaborators
Özgür Erkent and collaborators
Özgür Erkent and collaborators
Özgür Erkent and collaborators
E. Milli, Ö. Erkent and A. E. Yılmaz
M. Diaz-Zapata, D. Sierra-Gonzalez, Ö. Erkent, C. Laugier and J. Dibangoye
D. Sierra González, A. Paigwar, Ö. Erkent and C. Laugier
G. Salazar-Gomez et al.
M. Diaz-Zapata, Ö. Erkent, C. Laugier, J. Dibangoye and D. Sierra-Gonzalez