Semi-supervised framework for writer identification using structural learning

Writer identification is a complex task as the handwriting of an individual encapsulates lot of information pertaining to text and personality of a writer. To learn a model to distinguish one writer from the other, it is important to capture every nuance of the handwriting of an individual. Learning such model poses two challenges. First, discriminatory variables maybe large and potentially related leading to a complex discriminatory function. Second, it will require large amount of training data to learn a complex and possibly high-dimensional function. In this study, the authors are proposing a semi-supervised framework for writer identification for offline handwritten documents that leverages the information hidden in the unlabelled samples. Proposed framework models the complexity of approximating the optimal hypothesis by breaking the main task into several subtasks and learning a separate hypothesis for each subtask. All the hypotheses pertaining to the subtasks will be used for the best model selection by retrieving a common substructure that has high correspondence with all the candidate hypotheses. The obtained substructure acts as a knowledge base that has the contextual information, which is otherwise difficult to retrieve. The extra information can be used to improve the performance of the identification model.