Generalized Zero-Shot Learning via Synthesized Examples

TL;DR

This paper introduces a generative variational autoencoder framework for generalized zero-shot learning that synthesizes examples from class attributes, improving classification performance on both seen and unseen classes.

Contribution

The paper proposes a novel VAEs-based model with a feedback mechanism that generates exemplars for unseen classes, enhancing zero-shot learning capabilities.

Findings

Outperforms state-of-the-art methods on benchmark datasets

Effectively mitigates bias towards seen classes

Works for both standard and generalized zero-shot learning

Abstract

We present a generative framework for generalized zero-shot learning where the training and test classes are not necessarily disjoint. Built upon a variational autoencoder based architecture, consisting of a probabilistic encoder and a probabilistic conditional decoder, our model can generate novel exemplars from seen/unseen classes, given their respective class attributes. These exemplars can subsequently be used to train any off-the-shelf classification model. One of the key aspects of our encoder-decoder architecture is a feedback-driven mechanism in which a discriminator (a multivariate regressor) learns to map the generated exemplars to the corresponding class attribute vectors, leading to an improved generator. Our model's ability to generate and leverage examples from unseen classes to train the classification model naturally helps to mitigate the bias towards predicting seen classes in generalized zero-shot learning settings. Through a comprehensive set of experiments, we show that our model outperforms several state-of-the-art methods, on several benchmark datasets, for both standard as well as generalized zero-shot learning.