Postdoc in Robotics with Specialization in Visual Domain Adaptation

Postdoc in Robotics with Specialization in Visual Domain Adaptation

KTH - Royal Institute of Technology

Stockholm, Sweden

Job description

The Division of Robotics, Perception and Learning (RPL) at KTH Royal Institute of Technology invites applications for a postdoctoral researcher to lead an exciting project on visual domain adaptation for mobile robot perception, such as that of future delivery robots.

The focus of the project is on enabling mobiles robots to operate robustly in diverse real-world conditions, specifically tackling the challenge of transferring learned models from predominantly outdoor navigation to varied indoor environments for final delivery steps. A key research goal is to develop methods for fast and largely autonomous domain adaptation, with minimal human supervision. To this end, the project will explore the use of modern open-set vision models, the cloud computing infrastructure at Ericsson, as well as our robotics infrastructure, allowing for scalable, data-driven adaptation that can happen in hours rather than days. Experiments will be carried out with our Spot robot from Boston Dynamics, and there will be research engineer resources attached to the project to aid with these.

The postdoc will:

• Lead the research and development of domain adaptation methods for perception and navigation.
• Work in close collaboration with both academic researchers at RPL and industrial research team at Ericsson.
• Oversee experimental work on the Spot robot, with support from a part-time research engineer to facilitate practical deployments and data collection.

We are looking for candidates with a strong background in robotics, machine learning (preferably domain adaptation, sim2real, or representation learning), and experience with real robotic systems.

Qualifications

Requirements:

• A doctoral degree or an equivalent foreign degree. This eligibility requirement must be met no later than the time the employment decision is made.
• Very good programming skills in Python and using ML-frameworks (such as PyTorch).
• Strong expertise in machine learning applied to computer vision or robotic perception.
• Experience writing reusable and maintainable code for others to use.
• As a person, you have research expertise, independent and have the ability to collaborate.

Preferred qualifications:

• A doctoral degree or an equivalent foreign degree, obtained within the last three years prior to the application deadline
• Strong C++ skills.
• Practical experience working with robots.
• We prefer that you have pedagogical abilities.
• It is commendable if you Awareness of diversity and equal opportunity issues, with specific focus on gender equality.

Apply Now