An Entropy-guided Reinforced Partial Convolutional Network for Zero-Shot Learning

TL;DR

This paper introduces ERPCNet, a novel entropy-guided reinforced partial convolutional network that dynamically discovers and aggregates localities for improved zero-shot learning, achieving faster convergence and better performance without human annotations.

Contribution

ERPCNet is the first to combine entropy guidance with reinforced partial convolution for automatic locality discovery in ZSL, enhancing efficiency and interpretability.

Findings

ERPCNet outperforms state-of-the-art methods on four benchmark datasets.

It converges faster due to reinforcement learning and entropy guidance.

The model is time-efficient and provides explainable visualizations.

Abstract

Zero-Shot Learning (ZSL) aims to transfer learned knowledge from observed classes to unseen classes via semantic correlations. A promising strategy is to learn a global-local representation that incorporates global information with extra localities (i.e., small parts/regions of inputs). However, existing methods discover localities based on explicit features without digging into the inherent properties and relationships among regions. In this work, we propose a novel Entropy-guided Reinforced Partial Convolutional Network (ERPCNet), which extracts and aggregates localities progressively based on semantic relevance and visual correlations without human-annotated regions. ERPCNet uses reinforced partial convolution and entropy guidance; it not only discovers global-cooperative localities dynamically but also converges faster for policy gradient optimization. We conduct extensive experiments to demonstrate ERPCNet's performance through comparisons with state-of-the-art methods under ZSL and Generalized Zero-Shot Learning (GZSL) settings on four benchmark datasets. We also show ERPCNet is time efficient and explainable through visualization analysis.