Feature Generating Networks for Zero-Shot Learning
Authors:
- Yongqin XianYXYongqin Xian
- Tobias LorenzTLTobias Lorenz
- Bernt SchieleBSBernt Schiele
- Zeynep AkataZAZeynep Akata
Top Cited Papers
- 1 June 2018
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 5542-5551
Abstract
Suffering from the extreme training data imbalance between seen and unseen classes, most of existing state-of-the-art approaches fail to achieve satisfactory results for the challenging generalized zero-shot learning task. To circumvent the need for labeled examples of unseen classes, we propose a novel generative adversarial network (GAN) that synthesizes CNN features conditioned on class-level semantic information, offering a shortcut directly from a semantic descriptor of a class to a class-conditional feature distribution. Our proposed approach, pairing a Wasserstein GAN with a classification loss, is able to generate sufficiently discriminative CNN features to train softmax classifiers or any multimodal embedding method. Our experimental results demonstrate a significant boost in accuracy over the state of the art on five challenging datasets - CUB, FLO, SUN, AWA and ImageNet - in both the zero-shot learning and generalized zero-shot learning settings.Keywords
All Related Versions
This publication has 8 references indexed in Scilit:
- Low-Shot Visual Recognition by Shrinking and Hallucinating Features2017
- Photographic Image Synthesis with Cascaded Refinement Networks2017
- Learning a Deep Embedding Model for Zero-Shot Learning2017
- Zero-Shot Learning — The Good, the Bad and the Ugly2017
- Learning Deep Representations of Fine-Grained Visual Descriptions2016
- Learning Hypergraph-regularized Attribute Predictors2015
- SUN attribute database: Discovering, annotating, and recognizing scene attributes2012
- ImageNet: A large-scale hierarchical image database2009