Historically, we have long been interested in automatic face recognition, developing in particular new ways to find and describe the eyes in face images, and extending the feature vector for automatic face recognition. When we changed to moving feature extraction we were also interested in newer biometrics, especially gait and ears.
We are well known for our early work in automatic gait recognition, which is now where many of our students apply their new moving-feature extraction and description techniques. The advantages of using gait as a biometric are that it is non-contact and sequence based and has unique advantage at a distance when other biometrics cannot be used since they are at too-low resolution or obscured. We have support from Shakespeare too - he often describes the human capability to recognise people by the way they walk. Our work has recently created a considerable amount of media interest (in national and international TV, radio and the press).
Recently, we have been working on automatic ear recognition, where we develop a unique signature from the image of an ear (it has its own unique advantages). Even more recently we have moved to multibiometric data fusion, to use one biometric to reinforce the recognition by another.
We continue to work on automatic recognition by gait and on newer biometrics. These can give us ways to investigate our new feature extraction and description techniques, and in a technology of increasing international interest.