Summary
Aarhus University invites applications for a fully funded PhD fellowship within the Department of Electrical and Computer Engineering. The successful candidate will join the Adaptive & Agentic AI (A3) Lab, working in a unique interdisciplinary environment at the intersection of Foundation Models and Edge Intelligence. The project focuses on developing a high-performance, low-latency framework for Test-Time Adaptation (TTA) to ensure the reliability of AI models in volatile, real-world edge environments.
PhD Position in Efficient Test-Time Model Adaptation in Dynamic Edge Environments, Aarhus University, Denmark
Position Overview
| Detail | Information |
| Designation | PhD Fellow / Scholar |
| Host Institution | Graduate School of Technical Sciences, Aarhus University |
| Department | Electrical and Computer Engineering |
| Location | Adaptive & Agentic AI (A3) Lab, Finlandsgade 22, 8200 Aarhus N., and Denmark Research Centre Flakkebjerg, Forsøgsvej 1, DK-4200 Slagelse, Denmark |
| Preferred Starting Date | 01 November 2026 or later |
| Last Date to Apply | 15 August 2026 at 23:59 CEST |
Research Area
This doctoral research focuses on Foundation Models and Edge Intelligence. Traditional AI models often become brittle under distribution shifts caused by fluctuating conditions or hardware degradation at the edge. To avoid the latency penalties of cloud-based recalibration, this project aims to make edge AI systems “self-aware” and capable of autonomous evolution.
Job Description
The primary objective of this PhD is to design autonomous architectures capable of monitoring and maintaining the reliability of unimodal and multimodal foundation models in real-time. Key responsibilities and research pillars include:
- Autonomous Monitoring: Develop mechanisms to detect distribution shifts and quantify model uncertainty across heterogeneous data types.
- On-the-Fly Adaptation: Design lightweight TTA algorithms to recalibrate models at the edge under strict computational and latency constraints.
- Efficiency and Reliability: Balance the trade-offs between adaptation accuracy, energy efficiency, and hard real-time execution.
- Publication & Validation: Publish findings at top-tier machine learning venues (e.g., NeurIPS, ICLR, CVPR) and validate research on state-of-the-art edge computing testbeds.
Eligibility and Qualifications
Applicants must hold a master’s degree (120 ECTS) in Computer Science, Computer Engineering, Electrical Engineering, Machine Learning, or a closely related quantitative field.
Specific Competencies Required:
- Technical Skills: Advanced proficiency in Python and deep learning frameworks (e.g., PyTorch).
- Core Knowledge: Strong foundation in machine learning and/or computer vision, with specific interest in test-time adaptation, autonomous AI systems, and edge intelligence.
- Advanced Architectures & Edge AI: Familiarity with modern neural networks. Experience with edge-specific model compression (knowledge distillation, lightweight design, parameter-efficient fine-tuning) is highly advantageous.
- Mindset: Dedication to reproducibility, open-source contribution, and bridging algorithmic AI with practical edge systems.
How to Apply
To apply, please submit your application via this link before the deadline.
Mandatory Application Documents:
- Statement of Interest (1 page): Detail your background in ML/CV, relevant work experience, and your motivation for joining the A3 Lab.
- Curriculum Vitae: Include a publication list (if applicable) and a technical project portfolio.
- Academic Records: Transcripts and diplomas for both Bachelor’s and Master’s degrees.
- Project Description: For technical reasons, you must copy the project description from the announcement and upload it as a PDF.
For further information regarding mandatory attachments and requirements, please consult the application guide.
Contacts
- Main Supervisor: Behzad Bozorgtabar (behzad@ece.au.dk)
- Co-supervisor: Qi Zhang (qz@ece.au.dk)
- Admissions: admission.gradschool.tech@au.dk
Last Date for Apply: Open Now
