Transductive Zero-Shot Learning with a Self-training dictionary approach

TL;DR

This paper introduces a transductive zero-shot learning method using a self-training dictionary approach that models semantic relationships through bidirectional mappings and iterative refinement, improving recognition of unseen classes.

Contribution

It proposes a novel transductive ZSL framework with a self-training dictionary method that effectively addresses domain shift by iterative bootstrapping and bidirectional semantic modeling.

Findings

Outperforms state-of-the-art on AwA, CUB, and SUN datasets.

Effectively mitigates domain shift in zero-shot learning.

Demonstrates robustness through iterative self-training.

Abstract

As an important and challenging problem in computer vision, zero-shot learning (ZSL) aims at automatically recognizing the instances from unseen object classes without training data. To address this problem, ZSL is usually carried out in the following two aspects: 1) capturing the domain distribution connections between seen classes data and unseen classes data; and 2) modeling the semantic interactions between the image feature space and the label embedding space. Motivated by these observations, we propose a bidirectional mapping based semantic relationship modeling scheme that seeks for crossmodal knowledge transfer by simultaneously projecting the image features and label embeddings into a common latent space. Namely, we have a bidirectional connection relationship that takes place from the image feature space to the latent space as well as from the label embedding space to the latent space. To deal with the domain shift problem, we further present a transductive learning approach that formulates the class prediction problem in an iterative refining process, where the object classification capacity is progressively reinforced through bootstrapping-based model updating over highly reliable instances. Experimental results on three benchmark datasets (AwA, CUB and SUN) demonstrate the effectiveness of the proposed approach against the state-of-the-art approaches.