Fully-funded Research Assistant - Deep Semantic Models in Mobile Robotics


We are looking for an outstanding research assistant to join the development of deep learning solutions for semantic reasoning and planning in mobile robotics in a joint project between KTH Royal Institute of Technology in Stockholm and the University of Washington in Seattle. The position is full-time, fully funded for the period of up to 1 year, and open immediately. The position is hosted by KTH in Stockholm, the largest and oldest technical university in Sweden, providing one-third of Sweden's technical research and engineering education capacity at university level.

Successful candidates will work as members of a small, fast-paced research team focused on developing novel deep learning algorithms for multi-modal (color, depth) semantic perception, semantic environment understanding and action planning as well as their real-world deployment on mobile robots in large human-populated environments. The project is a collaboration between:

The research assistant will be undertaking both research and technical tasks, including:

  • Research related to novel deep learning techniques, such as Sum-Product Networks and Generative Adversarial Networks
  • Implementing deep learning models in TensorFlow
  • Contributing to a new, open-source library for inference and learning with Sum-Product Networks and Tensorflow
  • Deploying deep learning models on large-scale datasets and mobile robots
  • Developing components of an integrated robotics system based on ROS


  • Location: primarily KTH Royal Institute of Technology in Stockholm; possible visits at the University of Washington in Seattle
  • Salary: 26000SEK/month (funded by the Swedish Research Council)
  • Benefits: full package including health insurance and pension contributions
  • Duration: up to 1 year
  • Application deadline: immediately, candidates are encouraged to apply now
  • Expected start date: as soon as possible

Please direct all questions about the positions to:
Dr. Andrzej Pronobis



  • Eligibility to work in Sweden
  • M.S. degree or currently pursuing a PhD in CS or related fields
  • Very good knowledge of Linux and experience with development on Linux
  • Excellent, demonstrated coding skills in Python and C++
  • Experience with Python scientific libraries (Numpy, Scipy)
  • Familiarity with standard dev tools: shell scripting, CMake, gcc, gdb, git
  • Coursework and/or experience in at least one of the following areas: robotics, computer vision, pattern recognition, machine learning
  • Proven ability to design and implement scientific algorithms
  • Proficiency in English (oral and written), excellent communication skills


  • Experience with deep learning software frameworks, e.g., TensorFlow, Caffe
  • Experience contributing to open source projects, e.g. on GitHub
  • Experience contributing to research projects or scientific publications
  • Experience with developing robotic systems in ROS
  • Experience with computer vision libraries, e.g., OpenCV, PCL


To apply for the positions, please send:

  • Cover letter
  • CV/resume
  • Transcript
  • Links to previous projects and/or open-source contributions
  • Two references and/or two letters of recommendation
to Dr. Andrzej Pronobis (ude.notgnihsaw.sc@sibonorp). Please indicate the desired employment time frame. Eligibility to work in Sweden is essential.