Improved Feature Generating Framework for Transductive Zero-shot Learning

TL;DR

This paper introduces I-VAEGAN, a novel framework for transductive zero-shot learning that mitigates prior bias effects by using pseudo-conditional learning and variational embedding, achieving state-of-the-art results.

Contribution

The paper proposes I-VAEGAN, which incorporates PFA and VER to improve unseen class feature generation without prior estimation, advancing TZSL performance.

Findings

Achieves state-of-the-art TZSL accuracy on multiple benchmarks.

Effectively reduces prior bias impact in transductive ZSL.

Demonstrates robustness across various prior settings.

Abstract

Feature Generative Adversarial Networks have emerged as powerful generative models in producing high-quality representations of unseen classes within the scope of Zero-shot Learning (ZSL). This paper delves into the pivotal influence of unseen class priors within the framework of transductive ZSL (TZSL) and illuminates the finding that even a marginal prior bias can result in substantial accuracy declines. Our extensive analysis uncovers that this inefficacy fundamentally stems from the utilization of an unconditional unseen discriminator - a core component in existing TZSL. We further establish that the detrimental effects of this component are inevitable unless the generator perfectly fits class-specific distributions. Building on these insights, we introduce our Improved Feature Generation Framework, termed I-VAEGAN, which incorporates two novel components: Pseudo-conditional Feature Adversarial (PFA) learning and Variational Embedding Regression (VER). PFA circumvents the need for prior estimation by explicitly injecting the predicted semantics as pseudo conditions for unseen classes premised by precise semantic regression. Meanwhile, VER utilizes reconstructive pre-training to learn class statistics, obtaining better semantic regression. Our I-VAEGAN achieves state-of-the-art TZSL accuracy across various benchmarks and priors. Our code would be released upon acceptance.