Local descriptors to improve off-line handwriting-based gender prediction

Gender prediction based on the handwritten text becomes to earn a considerable importance for the document analysis community Gender prediction based on the handwritten text becomes to earn a considerable importance for the document analysis community. It is helpful for person identification as well as in some situations where one needs to classify population according to women-men categories. However, only a few studies have been carried out in this field. In the present work, we propose the use of local descriptors in order to improve the gender classification based on offline handwritten text. Specifically, we employ Histogram of Oriented Gradients (HOG), Local Binary Patterns (LBP) as well as grid features, which are successful in various pattern recognition applications. The prediction task is achieved by SVM classifier. The results obtained on samples extracted from IAM dataset show that local descriptors provide quite promising results.