Patch-based similarity HMMs for face recognition with a single reference image
Résumé
In this paper we present a new architecture for face recognition with a single reference image, which separates the training process from the recognition process completely. In the training stage, by using a database containing various individuals, the spatial relations between face components are represented by two Hidden Markov Models, one of which models the similarity between face images coming from the same individual (within-subject similarities), and the other modelling inter-subject differences. This allows us to take a pair of face images, neither of which has been seen before, and determine whether or not they come from the same individual. Whilst other face-recognition HMMs use Maximum Likelihood, we have tested our approach using both Maximum Likelihood and MAP estimation, and find that MAP estimation provides better results. Importantly, the training database can be entirely separated from the gallery and test images: this means that adding new individuals to the system can be done without re-training. We present results based upon models trained on the FERET training dataset, and demonstrate that these give excellent recognition rates on both the FERET database itself and more impressively the unseen AR database. When compared to other HMM based face recognition techniques, our algorithm is of much lower complexity due to the small size of our observation sequence.