Celebrating EEE's academic promotions: Professor Krystian Mikolajczyk

Developing intelligent systems that can "read" helps build a bigger picture of our world.

Six academic staff from our department have received promotions in the College's 2021 promotions round. Today we meet Professor Krystian Mikolajczyk.

Krystian Mikolajczyk, promoted to Professor in Computer Vision and Machine Learning

"My research activities are in computer vision and machine learning for visual recognition, for tasks including image and video classification, object recognition and tracking, visual surveillance, face recognition, medical imaging and image retrieval.

I joined the department in 2015, having worked in a number of academic research institutions across Europe, developing intelligent systems for applications ranging from autonomous robots, augmented reality, television archives, law enforcement, to animal conservation.

Modelling our 3D world

My research goal is developing systems that can understand visual information — one of the big challenges in AI. My main focus is on efficient methods for analysing visual data that can then be used in various AI applications. In particular matching and retrieval, which is a process of finding correspondences between different sets of data — multiple images or videos, data from different sensors, times, depths, or viewpoints. This is essential for building models of the 3D world reconstructed from 2D images and for augmented reality, which are used extensively by the film and TV industry.

We're also using this matching and 3D modelling method in Robotics for navigation and interaction with the environment. A robot simply picking up an object requires its 3D model to be efficiently produced in real time.

Machines that can 'read'

The semantic content of visual data — meaning content that can be read and understood by humans — is often described in natural language. An important area of my research interest is in building machines that can automatically “read” and reason about these visual and language inputs, and help generate better, more informative sentence-based image annotations which will improve our ability to search and browse huge collections of images. This was one of the research problems we have addressed in collaboration with the BBC for their television archives.

Among many projects I have been involved in, I particularly enjoyed developing camera-based technology for monitoring wildlife for conservation in Africa and Central America together with Zoological Society of London.

Imperial life

I lead a research group MatchLab where we extensively use machine learning and deep learning to advance fundamental research and engineer efficient systems for practical applications. I am also a member of the interdisciplinary research centre Intelligent Systems and Networks Group.

The highlights of my academic life at Imperial are the interactions with our excellent students and colleagues. I teach several undergraduate and MSc courses in the area of machine learning, computer vision and pattern recognition and I am a programme director of our specialist MSc in Applied Machine Learning."

Our congratulations to Krystian — and to all our staff on their recent promotions — from everyone in the EEE Community!