In this paper, we propose an automated framework for multi-view image classification tasks. The proposed framework is able to, all at once, train a model to find a common latent representation and perform data imputation, choose the best classifier and tune all necessary hyper-parameters.

Continue reading