Leveraging the Invariant Side of Generative Zero-Shot Learning

Abstract

Conventional zero-shot learning (ZSL) methods generally learn an embedding, e.g., visual-semantic mapping, to handle the unseen visual samples via an indirect manner. In this paper, we take the advantage of generative adversarial networks (GANs) and propose a novel method, named leveraging invariant side GAN (LisGAN), which can directly generate the unseen features from random noises which are conditioned by the semantic descriptions. Specifically, we train a conditional Wasserstein GANs in which the generator synthesizes fake unseen features from noises and the discriminator distinguishes the fake from real via a minimax game. Considering that one semantic description can correspond to various synthesized visual samples, and the semantic description, figuratively, is the soul of the generated features, we introduce soul samples as the invariant side of generative zero-shot learning in this paper. A soul sample is the meta-representation of one class. It visualizes the most semantically-meaningful aspects of each sample in the same category. We regularize that each generated sample (the varying side of generative ZSL) should be close to at least one soul sample (the invariant side) which has the same class label with it. At the zero-shot recognition stage, we propose to use two classifiers, which are deployed in a cascade way, to achieve a coarse-to-fine result. Experiments on five popular benchmarks verify that our proposed approach can outperform state-of-the-art methods with significant improvements.

Accepted in CVPR 2019

References in corpus (1)

• Conditional Generative Adversarial Nets

Cited by in corpus (22)

• Fine-Grained Zero-Shot Learning with DNA as Side Information
• Generative Adversarial Zero-shot Learning via Knowledge Graphs
• FREE: Feature Refinement for Generalized Zero-Shot Learning
• Generalized Continual Zero-Shot Learning
• Domain segmentation and adjustment for generalized zero-shot learning
• Hardness Sampling for Self-Training Based Transductive Zero-Shot Learning
• Zero-Shot Learning with Knowledge Enhanced Visual Semantic Embeddings
• Multi-Knowledge Fusion for New Feature Generation in Generalized Zero-Shot Learning
• GTNet: Generative Transfer Network for Zero-Shot Object Detection
• Rethinking Generative Zero-Shot Learning: An Ensemble Learning Perspective for Recognising Visual Patches
• Entropy-Based Uncertainty Calibration for Generalized Zero-Shot Learning
• Bi-Semantic Reconstructing Generative Network for Zero-shot Learning
• Structured Latent Embeddings for Recognizing Unseen Classes in Unseen Domains
• Improving Generalized Zero-Shot Learning by Semantic Discriminator
• OntoZSL: Ontology-enhanced Zero-shot Learning
• End-to-end Generative Zero-shot Learning via Few-shot Learning
• Bias-Awareness for Zero-Shot Learning the Seen and Unseen
• Towards Visual Explainable Active Learning for Zero-Shot Classification
• Zero-Shot Learning from Adversarial Feature Residual to Compact Visual Feature
• DFS: A Diverse Feature Synthesis Model for Generalized Zero-Shot Learning
• Mitigating Generation Shifts for Generalized Zero-Shot Learning
• Toward Zero-Shot Unsupervised Image-to-Image Translation