CIZSL++: Creativity Inspired Generative Zero-Shot Learning

Abstract

Zero-shot learning (ZSL) aims at understanding unseen categories with no training examples from class-level descriptions. To improve the discriminative power of ZSL, we model the visual learning process of unseen categories with inspiration from the psychology of human creativity for producing novel art. First, we propose CIZSL-v1 as a creativity inspired model for generative ZSL. We relate ZSL to human creativity by observing that ZSL is about recognizing the unseen, and creativity is about creating a likable unseen. We introduce a learning signal inspired by creativity literature that explores the unseen space with hallucinated class-descriptions and encourages careful deviation of their visual feature generations from seen classes while allowing knowledge transfer from seen to unseen classes. Second, CIZSL-v2 is proposed as an improved version of CIZSL-v1 for generative zero-shot learning. CIZSL-v2 consists of an investigation of additional inductive losses for unseen classes along with a semantic guided discriminator. Empirically, we show consistently that CIZSL losses can improve generative ZSL models on the challenging task of generalized ZSL from a noisy text on CUB and NABirds datasets. We also show the advantage of our approach to Attribute-based ZSL on AwA2, aPY, and SUN datasets. We also show that CIZSL-v2 has improved performance compared to CIZSL-v1.

arXiv commentThis paper is an extended version of a paper published on the International Conference on Computer Vision (ICCV), held in Seoul, Republic of Korea, October 27-Nov 2nd, 2019 CIZSL-v2 code is available here https://github.com/Vision-CAIR/CIZSLv2. arXiv admin note: substantial text overlap with arXiv:1904.01109

References in corpus (16)

• Conditional Image Synthesis With Auxiliary Classifier GANs
• Improved Training of Wasserstein GANs
• Zero-Shot Learning Through Cross-Modal Transfer
• Label-Embedding for Image Classification
• Zero-Shot Learning by Convex Combination of Semantic Embeddings
• A Learned Representation For Artistic Style
• A Neural Representation of Sketch Drawings
• Zero-Shot Learning -- A Comprehensive Evaluation of the Good, the Bad and the Ugly
• Learning Deep Representations of Fine-grained Visual Descriptions
• Semantic Autoencoder for Zero-Shot Learning
• From Zero-shot Learning to Conventional Supervised Classification: Unseen Visual Data Synthesis
• Latent Embedding Feedback and Discriminative Features for Zero-Shot Classification
• Is Sharma-Mittal entropy really a step beyond Tsallis and Renyi entropies?
• Multi-Cue Zero-Shot Learning with Strong Supervision
• Learning a Deep Embedding Model for Zero-Shot Learning
• Write a Classifier: Predicting Visual Classifiers from Unstructured Text