Develop graph-based inverse reinforcement learning and vision methods as a Postdoctoral Fellow with the NRC in Hamilton. Work full time onsite. This two-year fellowship focuses on driver attention strategies during adverse weather. You will build a graph-based inverse reinforcement learning framework for analyzing and forecasting human-driver eye-gaze scanpaths. The work combines spectral graph analysis, deep learning, and prediction of future gaze trajectories and driving actions. You will also participate in multidisciplinary research projects under direct supervision from an NRC researcher.

Key Responsibilities
- Develop graph-based inverse reinforcement learning frameworks
- Analyze and forecast driver eye-gaze scanpaths
- Apply spectral graph analysis to gaze patterns




- Forecast future gaze trajectories and driving actions
- Participate in multidisciplinary research projects

Requirements:
- PhD in Computer Engineering, Computer Science, or related field
- Two to six years developing advanced deep learning architectures
- Experience with transformer models, CNNs, generative, diffusion, and reinforcement learning models
- Clear, concise coding experience in Python or C++
- Experience with PyTorch, TensorFlow, Scikit-learn, NumPy, Pandas, and OpenCV Apply deep learning, graph analysis, and computer vision research to driver attention forecasting at the NRC in Hamilton.

📌 Postdoctoral Fellow Reinforcement Learning Vision (Canada)
🏢 National Research Council of Canada
📍 Canada