PhD in Perception for Dynamic Scene Understanding, 3D Vision, and Autonomous Driving KTH Royal Institute of Technology in Sweden

A PhD position is open at KTH Royal Institute of Technology in the area of perception for dynamic scene understanding, with a strong focus on 3D vision, scene flow estimation, robotics, machine learning, and autonomous driving.

The project is based in the Department of Robotics, Perception and Learning and aims to develop a self-supervised, multi-modal model that combines LiDAR geometry and camera semantics to learn motion representations that generalize to unseen real-world scenarios. The work is connected to the group’s prior research on self-supervised scene flow and is intended to support autonomous planning and safety-critical perception.

The position is funded by the Wallenberg AI, Autonomous Systems and Software Program (WASP), and the student will be connected to the WASP graduate school. The announcement mentions access to strong compute resources through NSC and a research environment with an active community and colleagues.

Eligibility: applicants must have basic eligibility for doctoral studies, typically a master’s degree or equivalent, and must meet the English requirement equivalent to English B/6. The selection process also values the ability to work independently and collaboratively, a professional approach, strong English communication, and experience programming and implementing machine learning methods, preferably in computer vision.

Application: submit the application through KTH’s recruitment system with diplomas, transcripts, proof of language requirements, CV, a motivation letter, and publications/technical reports. The deadline is 2026-05-15.