Zero-shot Learning (ZSL), which aims to predict for those classes that have never appeared in the training data, has arisen hot research interests. The key of implementing ZSL is to leverage the prior knowledge of classes which builds the semantic relationship between classes and enables the transfer of the learned models (e.g., features) from training classes (i.e., seen classes) to unseen classes. However, the priors adopted by the existing methods are relatively limited with incomplete semantics. In this paper, we explore richer and more competitive prior knowledge to model the inter-class relationship for ZSL via ontology-based knowledge representation and semantic embedding. Meanwhile, to address the data imbalance between seen classes and unseen classes, we developed a generative ZSL framework with Generative Adversarial Networks (GANs). Our main findings include: (i) an ontology-enhanced ZSL framework that can be applied to different domains, such as image classification (IMGC) and knowledge graph completion (KGC); (ii) a comprehensive evaluation with multiple zero-shot datasets from different domains, where our method often achieves better performance than the state-of-the-art models. In particular, on four representative ZSL baselines of IMGC, the ontology-based class semantics outperform the previous priors e.g., the word embeddings of classes by an average of 12.4 accuracy points in the standard ZSL across two example datasets (see Figure 4).
翻译:零点学习(ZSL)旨在预测那些在培训数据中从未出现过的班级(ZSL)的预测,它引起了激烈的研究兴趣。实施 ZSL的关键是利用以前班级的知识,这些班级在班级之间建立语义关系,使学习模型(例如特征)从训练班(即被看见的班级)转移到看不见班级。然而,现有方法采用的前期方法相对有限,语义不完整。在本文中,我们探索了更丰富和更具竞争力的先前知识,通过基于内科的知识表达和语义嵌入来模拟ZSL的班际关系。与此同时,为了解决所见班级和看不见班级之间的数据不平衡,我们开发了一个具有基因化的ZSL框架(例如特征)从训练班(即被看见的班级)转到看不见的班级。我们的主要调查结果包括:(一) 一个基于本的ZSLSL框架,可以适用于不同领域,例如图像分类(IMGC)和知识图表完成(KGC), (二) 以多个零点的SLSL- sal sal sal sal devely sli laveal laveal sal sal sal sal sal ex sal ex sal a pal ex ex ex ex laveal a pas ex ex ex ex ex a pas ex a ex ex a pal a ex a ex a ex a ex ex ex ex ex ex ex ex ex ex ex ex laut the pal sal sal sal sal sal sal sal a ex ex. ex lautes. a lements sal salds sal sal. a pal a pal sal. a lats. a pal. a lats.