PhD Studentship: Developing Human Gaze Estimation and Following in Natural Scenes

Applications are invited for a funded 3 year PhD studentship in Computer Vision and Deep Learning. This project will develop state-of-the-art algorithms and software solutions in exploiting and making advances in computer vision and learning techniques to move toward intelligent interaction with visual data. In particular, the latest advances in computer vision and deep learning will be explored significantly to recognise various attributes of human, e.g. eye gaze, body pose, hand posture, action, and intention.

During the project, you will develop advanced skills in computer vision, deep learning, and mathematical modelling, and will also gain an understanding of the practical aspects of human-robot interaction. In this project, you will focus on human eyes and their movements and work towards estimating gaze and beyond from visual data. There have been many approaches to utilise gaze for an important functional component in various applications, as it indicates human attentiveness and can thus be used to study their intentions and understand social interactions. Wearable eye tracker can detect accurate gaze and monitor eye movements, but it restricts natural human behaviours and movement ranges. For these reasons, accurately estimating and following gaze from appearance (not wearing an eye tracker) has been an active research topic in computer vision, with applications in affect analysis, saliency detection and action recognition to name a few. Within the robotics community, gaze estimation and following allows a robot to detect which object is manipulated by the human, so the robot can act accordingly.

The goal of this research is therefore to create a new appearance-based gaze reading system that overcomes existing methods only using visual data captured by generic RGB cameras such as a digital camcorder, webcam, smartphone camera, and CCTV camera etc. The result will be a significant human understanding capability that is relevant to many academic (especially cognitive science, psychology and sports science) and industrial (commercial industry, healthcare, and autonomous system) sectors. The successful candidate will join the Intelligent Robotics Lab at the School of Computer Science, University of Birmingham under the supervision of Dr Hyung Jin Chang (https://www.cs.bham.ac.uk/~changhj ).

Eligibility: First or Upper Second Class Honours undergraduate degree and/or postgraduate degree with Distinction (or an international equivalent) in Computer Science, Electronics, or closely related fields. Applicants also need to have a strong background in deep learning and high proficiency in programming (especially in Python). An MSc project in machine learning, deep learning, or computer vision related areas would be beneficial but not essential. We also consider applicants from diverse backgrounds that have provided them with equally rich relevant experience and knowledge. Full-time and part-time study modes are available. If your first language is not English and you have not studied in an English-speaking country, you will have to provide an English language qualification. Anticipated start date: 6 February 2023. Funding is through a Korea Government’s IITP research funding.