Role Description
This is a full-time postdoctoral position based at the University of Waterloo (Waterloo, Ontario), with on-site collaboration at EAIGLE’s office in the Greater Toronto Area once every two weeks.
The fellow will research world models that track and understand transport vehicles across multiple cameras and over time at logistics gates and yards, under highly uncertain operating conditions such as occlusion, poor weather, night, glare and long gaps between observations. The research covers:
- Spatio-temporal tracking of vehicles across multiple cameras and extended time periods
- Vehicle text recognition: license plates, trailer and container IDs, and other markings
- Vehicle type classification: tractors, trailers, tankers, containers and chassis
- Physical condition assessment: damage, doors, seals and changes over time
- Reasoning under uncertainty: fusing evidence across cameras and time into a consistent, confidence-aware picture of each vehicle
The fellow will develop and validate computer vision and machine learning models using large volumes of real-world data from EAIGLE deployments, along with synthetic data from EAIGLE’s simulation pipelines. They will run experiments, publish at top academic venues, and work with EAIGLE’s engineering and product teams to turn research into deployed solutions. The fellow will be supervised by faculty at the University of Waterloo, present findings to academic and industry stakeholders, mentor graduate students, and help shape future research directions and grant proposals.
Qualifications
- PhD in Computer Science, Electrical/Computer Engineering or a closely related field (completed or near completion)
- Strong publication record in computer vision, machine learning or robotics
- Experience in one or more of: multi-object or multi-camera tracking, world models, spatio-temporal modelling, scene text recognition/OCR, fine-grained recognition, or uncertainty estimation
- Hands-on experience with deep learning frameworks (PyTorch preferred) and computer vision libraries such as OpenCV
- Strong programming skills in Python (C++ an asset), with an emphasis on reproducible research and good software engineering practice
- Experience handling large-scale datasets and rigorously evaluating model performance
- Experience with synthetic data, simulation or sim-to-real transfer is an asset
- Mentoring or teaching experience, such as supervising students or leading tutorials
- Strong scholarly writing and communication skills, and the ability to collaborate across academic and industry teams
- Ability to work on-site at the University of Waterloo and travel to EAIGLE’s GTA office once every two weeks
