My research interest is in visual SLAM (simultaneous localisation and mapping) which is the general problem faced by any robot or other devices with the need to localise, map and navigate a space it has never been in before by utilising various cameras as the main outward looking sensors to achieve this. Although the technology has remarkably matured over the past few years, we still face several challenges: especially the response to rapid motion, handling of extreme lighting variation, minimising overall processing latency and power consumption. My research is therefore focused on developing more robust and efficient SLAM systems by utilising novel sensor and processor hardware technologies such as the event camera.

I have worked in various computer vision related fields including surveillance camera, sensor guided robot, automatic number plate recognition and visual SLAM. And, in my PhD, supervised by Prof. Andrew Davison at Imperial College London, we have developed the first methods which enable incremental reconstruction, tracking and SLAM from event camera data all based on rigorous probabilistic principles, and won the best industry paper award at the British Machine Vision Conference (BMVC) 2014 and the best paper award at the European Conference on Computer Vision (ECCV) 2016. I was also selected as one of the UK winners and the European champion of the Qualcomm Innovation Fellowship 2014 for my research proposal “Visual SLAM