Morphological Shared Weight Neural Network: A method to improve fault tolerance in Face Recognition

In this paper, the focus is on the investigation offace recognition using morphological shared-weight neuralnetwork. Being nonlinear and translation-invariant, the MSNNcan be used to create better generalization during facerecognition. The MSNN is a heterogeneous network that produceshigh order features based on local features extracted bymorphological operations. Feature extraction is performed ongrayscale images using hit-miss transforms that are independentof gray-level shifts. The output is then learned by interacting withthe classification process. The feature extraction andclassification networks are trained together, allowing the MSNNto simultaneously learn feature extraction and classification for aface.