Zero-shot Recognition via Semantic Embeddings and Knowledge Graphs

TL;DR

This paper introduces a method for zero-shot recognition that leverages semantic embeddings and knowledge graphs with graph convolutional networks to predict classifiers for unseen categories, significantly improving accuracy.

Contribution

It proposes a novel GCN-based approach that uses semantic embeddings and categorical relationships to predict classifiers for unseen classes in zero-shot recognition.

Findings

Achieves up to 20% accuracy on some metrics for unseen categories.

Demonstrates robustness to noise in the knowledge graph.

Significantly outperforms previous state-of-the-art methods.

Abstract

We consider the problem of zero-shot recognition: learning a visual classifier for a category with zero training examples, just using the word embedding of the category and its relationship to other categories, which visual data are provided. The key to dealing with the unfamiliar or novel category is to transfer knowledge obtained from familiar classes to describe the unfamiliar class. In this paper, we build upon the recently introduced Graph Convolutional Network (GCN) and propose an approach that uses both semantic embeddings and the categorical relationships to predict the classifiers. Given a learned knowledge graph (KG), our approach takes as input semantic embeddings for each node (representing visual category). After a series of graph convolutions, we predict the visual classifier for each category. During training, the visual classifiers for a few categories are given to learn the GCN parameters. At test time, these filters are used to predict the visual classifiers of unseen categories. We show that our approach is robust to noise in the KG. More importantly, our approach provides significant improvement in performance compared to the current state-of-the-art results (from 2 ~ 3% on some metrics to whopping 20% on a few).