PhD Position in Foundation Models for Automotive Imaging Radar
Develop next-generation foundation models for automotive radar perception to enable advanced 3D scene understanding and multimodal sensor fusion.
Best for:Best for computer science or robotics graduates with a strong background in machine learning, deep learning, and Python/PyTorch, interested in foundation models, self-supervised learning, and automotive sensor technology.
Overview
This PhD position at Delft University of Technology focuses on advancing foundation models for automotive imaging radar systems. The project aims to create generalizable radar representations from large-scale, largely unlabeled data to improve tasks like 3D object detection and free space estimation. The work involves methodological machine learning research, experiments with real automotive sensor data, and collaboration with industry partners Perciv.AI and NXP Semiconductors. The successful candidate will contribute to the FIND project, funded by the Dutch Research Council (NWO), and publish results in top venues such as NeurIPS, CVPR, and ICRA. Access to advanced computing resources and research vehicles will support the research.
What you'll do
- Develop and investigate foundation models for automotive imaging radar to enable generalizable 3D scene understanding and perception tasks.
- Design and implement self-supervised and multimodal learning approaches for radar data, leveraging insights from vision and LiDAR foundation models.
- Conduct experiments using real automotive sensor data, including radar prototypes and research vehicles, to validate and refine models.
- Collaborate with industry partners Perciv.AI and NXP Semiconductors to ensure practical applicability of research outcomes.
- Publish findings in leading international venues such as NeurIPS, CVPR, ICRA, IEEE IV, and RadarConf.
- Utilize advanced computing resources, including Delft AI Cluster and DelftBlue supercomputer, for large-scale model training and evaluation.
Requirements
Degree required: Master of Science (MSc) in Computer Science, Artificial Intelligence, Robotics, Electrical Engineering, or a closely related field
Degree offered: PhD
Language: English; proficiency required for communication, research, and publication
Instruction: English
- Strong academic record and solid background in machine learning and deep learning
- Proficiency in developing, understanding, and critically evaluating machine learning research software using Python and PyTorch
- Interest and experience in foundation models, self-supervised learning, multimodal learning, and 3D perception
- Ability to translate methodological research into experiments with real automotive sensor data and vehicle demonstrators
- Strong written and spoken English proficiency
- Experience with radar, signal processing, computer vision, autonomous driving, or sensor fusion
- No specific nationality or residency restrictions mentioned
- No age or prior experience caps mentioned
Funding & terms
How to apply
Deadline: 2026-10-18
Apply via the TU Delft application portal by uploading a motivation letter, detailed CV, BSc and MSc transcripts, supporting materials (e.g., thesis draft, publications, code repositories), and reference contact details upon request. Applications sent by email or post will not be processed.
- Motivation letter
- Detailed CV
- BSc and MSc transcripts (courses and grades)
- Supporting materials (e.g., link to MS Thesis draft, publications, project descriptions, code repositories)
- Reference contact details (provided upon interview selection)
Contact
Prof. Dr. D.M. Gavrila· Supervisor
Create a free profile to unlock your personal match score, the full requirements, funding details and how to apply — in English or Persian.