Recovering the Missing Link: Predicting Class-Attribute Associations for Unsupervised Zero-Shot Learning
This addresses the time-consuming and domain-specific need for manual attribute labeling in zero-shot learning, offering a more practical solution for image classification tasks.
The paper tackles the problem of requiring manual class-attribute associations for zero-shot learning by proposing an unsupervised method to predict these associations from class names, achieving significant performance improvements over state-of-the-art methods on datasets like Animals with Attributes and aPascal/aYahoo.
Collecting training images for all visual categories is not only expensive but also impractical. Zero-shot learning (ZSL), especially using attributes, offers a pragmatic solution to this problem. However, at test time most attribute-based methods require a full description of attribute associations for each unseen class. Providing these associations is time consuming and often requires domain specific knowledge. In this work, we aim to carry out attribute-based zero-shot classification in an unsupervised manner. We propose an approach to learn relations that couples class embeddings with their corresponding attributes. Given only the name of an unseen class, the learned relationship model is used to automatically predict the class-attribute associations. Furthermore, our model facilitates transferring attributes across data sets without additional effort. Integrating knowledge from multiple sources results in a significant additional improvement in performance. We evaluate on two public data sets: Animals with Attributes and aPascal/aYahoo. Our approach outperforms state-of-the-art methods in both predicting class-attribute associations and unsupervised ZSL by a large margin.