Context-Aware Zero-Shot Learning for Object Recognition
This work addresses a limitation in zero-shot learning for computer vision by modeling object-context relationships, offering incremental improvements for object recognition tasks.
The paper tackles the problem of zero-shot learning for object recognition by incorporating contextual information, which previous methods ignored, and shows that this approach substantially improves performance and robustness to unbalanced classes in experiments on Visual Genome.
Zero-Shot Learning (ZSL) aims at classifying unlabeled objects by leveraging auxiliary knowledge, such as semantic representations. A limitation of previous approaches is that only intrinsic properties of objects, e.g. their visual appearance, are taken into account while their context, e.g. the surrounding objects in the image, is ignored. Following the intuitive principle that objects tend to be found in certain contexts but not others, we propose a new and challenging approach, context-aware ZSL, that leverages semantic representations in a new way to model the conditional likelihood of an object to appear in a given context. Finally, through extensive experiments conducted on Visual Genome, we show that contextual information can substantially improve the standard ZSL approach and is robust to unbalanced classes.