PhD fellowship

PhD position in Efficient Test-Time Model Adaptation in Dynamic Edge Environments at Aarhus University, Denmark, in the Department of Electrical and Computer Engineering and the Adaptive & Agentic AI (A3) Lab.

The project sits at the intersection of foundation models, edge intelligence, machine learning, computer vision, and real-time adaptive AI. The research aims to build high-performance, low-latency test-time adaptation methods for unimodal and multimodal models operating in dynamic edge environments where data streams face domain shifts, hardware degradation, and changing physical conditions.

The successful candidate will work under Dr. Behzad Bozorgtabar (main supervisor) and Prof. Qi Zhang (co-supervisor). The project emphasizes autonomous monitoring, uncertainty estimation, on-the-fly adaptation, efficiency, and reliability for mission-critical applications such as autonomous robotics and industrial monitoring.

Funding: fully funded PhD fellowship/scholarship. Salary and employment terms follow the applicable collective agreement. No stipend amount is specified.

Eligibility: applicants must hold a master’s degree (120 ECTS) in Computer Science, Computer Engineering, Electrical Engineering, Machine Learning, or a related quantitative field. Strong Python and PyTorch skills are expected, together with a solid background in machine learning and/or computer vision. Experience with modern neural networks and edge-specific model compression techniques is advantageous.

Application materials: statement of interest (1 page), CV including publication list and technical portfolio, academic transcripts and diplomas, and a project description copied from the announcement and uploaded as a PDF.

Deadline: 15 August 2026 at 23:59 CEST. Preferred start date is 01 November 2026.

Applicants can submit via the official application portal linked in the announcement. For questions, contact [email protected] or [email protected].