A Generative Model For Zero Shot Learning Using Conditional Variational Autoencoders
This addresses the problem of classifying images from novel classes without training data, which is important for real-world applications where not all classes are available during training.
The paper tackles zero-shot learning in image classification by generating samples from class attributes using a conditional variational autoencoder, and shows that this approach outperforms state-of-the-art methods on four benchmark datasets, especially in generalized settings.
Zero shot learning in Image Classification refers to the setting where images from some novel classes are absent in the training data but other information such as natural language descriptions or attribute vectors of the classes are available. This setting is important in the real world since one may not be able to obtain images of all the possible classes at training. While previous approaches have tried to model the relationship between the class attribute space and the image space via some kind of a transfer function in order to model the image space correspondingly to an unseen class, we take a different approach and try to generate the samples from the given attributes, using a conditional variational autoencoder, and use the generated samples for classification of the unseen classes. By extensive testing on four benchmark datasets, we show that our model outperforms the state of the art, particularly in the more realistic generalized setting, where the training classes can also appear at the test time along with the novel classes.