Semantically Aligned Bias Reducing Zero Shot Learning
This work addresses key challenges in zero-shot learning for computer vision, offering incremental improvements over existing methods.
The paper tackles the hubness problem and bias towards seen classes in zero-shot learning by proposing SABR-ZSL, which learns a latent space preserving semantic relationships and reduces bias through cross-validation and weak transfer constraints, achieving performance improvements of ~1.5-9% in conventional ZSL and ~2-14% in generalized ZSL.
Zero shot learning (ZSL) aims to recognize unseen classes by exploiting semantic relationships between seen and unseen classes. Two major problems faced by ZSL algorithms are the hubness problem and the bias towards the seen classes. Existing ZSL methods focus on only one of these problems in the conventional and generalized ZSL setting. In this work, we propose a novel approach, Semantically Aligned Bias Reducing (SABR) ZSL, which focuses on solving both the problems. It overcomes the hubness problem by learning a latent space that preserves the semantic relationship between the labels while encoding the discriminating information about the classes. Further, we also propose ways to reduce the bias of the seen classes through a simple cross-validation process in the inductive setting and a novel weak transfer constraint in the transductive setting. Extensive experiments on three benchmark datasets suggest that the proposed model significantly outperforms existing state-of-the-art algorithms by ~1.5-9% in the conventional ZSL setting and by ~2-14% in the generalized ZSL for both the inductive and transductive settings.