Abstract

We present a novel latent embedding model for learning a compatibility function between image and class embed-dings, in the context of zero-shot classification. The proposed method augments the state-of-the-art bilinear compatibility model by incorporating latent variables. Insteadof learning a single bilinear map, it learns a collection of maps with the selection, of which map to use, being a la-tent variable for the current image-class pair. We train themodel with a ranking based objective function which penalizes incorrect rankings of the true class for a given image. We empirically demonstrate that our model improvesthe state-of-the-art for various class embeddings consistently on three challenging publicly available datasets for the zero-shot setting. Moreover, our method leads to visually highly interpretable results with clear clusters of different fine-grained object properties that correspond to different latent variable maps.