Dynamic VAEs with Generative Replay for Continual Zero-shot Learning

TL;DR

This paper introduces DVGR-CZSL, a novel continual zero-shot learning model that uses generative replay to prevent forgetting and effectively classifies unseen objects in sequential tasks.

Contribution

The paper proposes a growing model with generative replay for continual zero-shot learning, addressing catastrophic forgetting and improving classification of unseen classes.

Findings

DVGR-CZSL outperforms baseline models on multiple datasets.

The model effectively prevents catastrophic forgetting.

Demonstrates superior performance in sequential zero-shot learning tasks.

Abstract

Continual zero-shot learning(CZSL) is a new domain to classify objects sequentially the model has not seen during training. It is more suitable than zero-shot and continual learning approaches in real-case scenarios when data may come continually with only attributes for a few classes and attributes and features for other classes. Continual learning(CL) suffers from catastrophic forgetting, and zero-shot learning(ZSL) models cannot classify objects like state-of-the-art supervised classifiers due to lack of actual data(or features) during training. This paper proposes a novel continual zero-shot learning (DVGR-CZSL) model that grows in size with each task and uses generative replay to update itself with previously learned classes to avoid forgetting. We demonstrate our hybrid model(DVGR-CZSL) outperforms the baselines and is effective on several datasets, i.e., CUB, AWA1, AWA2, and aPY. We show our method is superior in task sequentially learning with ZSL(Zero-Shot Learning). We also discuss our results on the SUN dataset.