A New Framework for Semisupervised and Multitask Learning

Labeling image collections is a tedious and time consuming task, especially when multiple labels have to be chosen for each image. On the other hand, the explosion of Internet content has provided cheap access to almost unlimited amounts of data, albeit with a lower quality of annotations. This dissertation deals with the problem of automatically annotating images, by introducing a new framework that extends state-of-the-art models in word prediction to incorporate information from two sources, unlabeled examples and correlated labels. This is the first semisupervised multitask model used in vision problems of these characteristics.