Abstract
We propose a deep learning framework for image set classifification with application to face recognition. An Adaptive Deep Network Template (ADNT) is defifined whose parameters are initialized by performing unsupervised pretraining in a layer-wise fashion using Gaussian Restricted Boltzmann Machines (GRBMs). The pre-initialized ADNT is then separately trained for images of each class and class-specifific models are learnt. Based on the minimum reconstruction error from the learnt class-specifific models, a majority voting strategy is used for classifification. The proposed framework is extensively evaluated for the task of image set classifification based face recognition on Honda/UCSD, CMU Mobo, YouTube Celebrities and a Kinect dataset. Our experimental results and comparisons with existing state-of-the-art methods show that the proposed method consistently achieves the best performance on all these datasets