PhD in Embedded Software for Dynamic Neural Networks

KU Leuven; Sint-Katelijne-Waver; Sur place; EUR 30 000 - 42 000; Plein temps

Avantages offerts par ce poste

PhD scholarship (up to 4 years)

International conference participation

Commute cost reimbursement

Résumé du poste

KU Leuven invites a motivated PhD researcher to join the Embedded Systems unit of EAVISE for a project on dynamic embedded AI. The role centers on enabling devices to adapt neural networks on-device after deployment, with emphasis on run-time reconfigurable kernels and energy-efficient computing for next-generation wearables and IoT platforms.

You will collaborate across hardware-software co-design, compiler/runtime systems, and validation on state-of-the-art embedded AI hardware, contributing

Qualifications

• Master's degree in electrical or computer engineering or equivalent.
• Experience programming in C/C++ in embedded context preferred.
• Strong understanding of computer architectures and embedded platforms.

Responsabilités

• Conduct research on runtime-reconfigurable neural network software for embedded devices.
• Develop performance models for adaptive AI workloads.
• Explore on-device search and optimization for memory-aware deployment.

Connaissances

C/C++ programming

Embedded software

English proficiency

Formation

Master's degree in electrical or computer engineering

Outils

ARM Cortex-M

FPGA

NPU

Description du poste

PhD in Embedded Software for Dynamic Neural Networks

The Embedded Systems unit of EAVISE at KU Leuven is looking for a highly motivated PhD researcher to join a novel research initiative investigating how embedded AI systems can dynamically optimize themselves after deployment. EAVISE is a multidisciplinary research group developing state-of-the-art AI, embedded computing, computer vision, and edge intelligence solutions for real-world applications. The Embedded Systems unit focuses on hardware-software co-design, AI deployment tools, and energy-efficient computing for future intelligent devices.

Project

Modern embedded AI systems rely on Deep Neural Networks (DNNs) running on resource-constrained devices such as wearables, smart sensors, hearables, and IoT nodes. While current deployment methodologies can optimize models before deployment, the resulting software remains static throughout the lifetime of the device. Consequently, energy consumption, memory usage, and computational requirements remain fixed, even when the application or environmental conditions evolve over time.

This project investigates a radically new paradigm for Embedded AI: enabling devices to dynamically compress and adapt neural networks directly on-device after deployment. Inspired by how humans continuously optimize their physical efficiency through training, this project aims to develop software methods that allow deployed neural networks to progressively reduce their resource footprint while maintaining application performance.

The research focuses include: runtime-reconfigurable neural network software kernels for novel embedded systems architecture, performance models for adaptive AI workloads, on-device search and optimization engines for memory-aware deployment, compiler and runtime systems capable of supporting dynamic neural network topologies, energy-efficient embedded AI frameworks targeting RISC-V multicore and heterogeneous computing platforms.

The project combines fundamental research in embedded software and compiler systems with practical validation on state-of-the-art embedded AI hardware platforms. Successful outcomes could significantly extend battery lifetime and enable a new generation of adaptive and sustainable embedded AI systems

Profile

We are looking for highly motivated PhD researchers with a strong interest in hardware-software design for embedded systems and Embedded AI.

• Candidates must hold a master’s degree in electrical or computer engineering (or equivalent). Master's students who expect to graduate by the end of the year are also encouraged to apply.
• Programming experience (C/C++) within the embedded field and/or software engineering and compilers is highly recommended.
• Strong background in computer architectures and embedded platforms (ARM Cortex-M, NPU, FPGA, embedded GPU), e.g., via academic courses and/or project courses
• Research experience (e.g., through a Master thesis work or research internships) is considered a strong asset.
• Experience with Deep Learning and Artificial Intelligence is considered a plus.
• Excellent proficiency in the English language is required, as well as good communication skills, both oral and written.
• Strong interpersonal skills and the ability to work in an international team.

Offer

• A PhD scholarship for up to 4 years (subject to positive intermediate evaluations).
• An exciting research environment, working on the intersection between theory and implementation in a very multi-disciplinary research environment.
• A thorough scientific education in the frame of a doctoral training program, with the possibility of becoming a world-class researcher.
• A competitive salary or tax-free PhD grant, including the reimbursement of commute costs to the workplace.
• Flexible working conditions, upon agreement with the PhD supervisor
• The possibility to participate in international conferences, workshops, and collaborations with top EU and KU Leuven research teams.

KU Leuven strives for an inclusive, respectful and socially safe environment. We embrace diversity among individuals and groups as an asset. Open dialogue and differences in perspective are essential for an ambitious research and educational environment. In our commitment to equal opportunity, we recognize the consequences of historical inequalities. We do not accept any form of discrimination based on, but not limited to, gender identity and expression, sexual orientation, age, ethnic or national background, skin colour, religious and philosophical diversity, neurodivergence, employment disability, health, or socioeconomic status. For questions about accessibility or support offered, we are happy to assist you at this email address.