Feature Generating Networks for Zero-Shot Learning (2017-12-04T00:00:00.000000Z)

TL;DR

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.

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.

Authors

B. Schiele

35 papers

Yongqin Xian

6 papers

Zeynep Akata

19 papers

TL;DR

Abstract

Authors

References50 items

Generating Visual Representations for Zero-Shot Classification

Photographic Image Synthesis with Cascaded Refinement Networks

Improved Training of Wasserstein GANs

Zero-Shot Learning — The Good, the Bad and the Ugly

Wasserstein GAN

Towards Principled Methods for Training Generative Adversarial Networks

StackGAN: Text to Photo-Realistic Image Synthesis with Stacked Generative Adversarial Networks

Learning a Deep Embedding Model for Zero-Shot Learning

Learning What and Where to Draw

Improving Semantic Embedding Consistency by Metric Learning for Zero-Shot Classiffication

InfoGAN: Interpretable Representation Learning by Information Maximizing Generative Adversarial Nets

Improved Techniques for Training GANs

Low-Shot Visual Recognition by Shrinking and Hallucinating Features

Learning Deep Representations of Fine-Grained Visual Descriptions

Generative Adversarial Text to Image Synthesis

An Empirical Study and Analysis of Generalized Zero-Shot Learning for Object Recognition in the Wild

Semi-supervised Vocabulary-Informed Learning

Less is More: Zero-Shot Learning from Online Textual Documents with Noise Suppression

Multi-cue Zero-Shot Learning with Strong Supervision

Latent Embeddings for Zero-Shot Classification

Generative Image Modeling Using Style and Structure Adversarial Networks

Synthesized Classifiers for Zero-Shot Learning

Deep Residual Learning for Image Recognition

Semi-Supervised Zero-Shot Classification with Label Representation Learning

Unsupervised Domain Adaptation for Zero-Shot Learning

Unsupervised Representation Learning with Deep Convolutional Generative Adversarial Networks

Zero-Shot Learning via Semantic Similarity Embedding

An embarrassingly simple approach to zero-shot learning

Zero-shot object recognition by semantic manifold distance

Label-Embedding for Image Classification

Learning Hypergraph-regularized Attribute Predictors

Generative Moment Matching Networks

Transductive Multi-View Zero-Shot Learning

Conditional Generative Adversarial Nets

Evaluation of output embeddings for fine-grained image classification

Zero-Shot Learning by Convex Combination of Semantic Embeddings

Transfer Learning in a Transductive Setting

DeViSE: A Deep Visual-Semantic Embedding Model

Distributed Representations of Words and Phrases and their Compositionality

Zero-Shot Learning Through Cross-Modal Transfer

SUN attribute database: Discovering, annotating, and recognizing scene attributes

Evaluating knowledge transfer and zero-shot learning in a large-scale setting

Caltech-UCSD Birds 200

Attribute-Based Transfer Learning for Object Categorization with Zero/One Training Example

Automated Flower Classification over a Large Number of Classes

Zero-data Learning of New Tasks

GENERATIVE ADVERSARIAL NETS

SMOTE: Synthetic Minority Over-sampling Technique

Attribute-Based Classification for Zero-Shot Visual Object Categorization

and as an in

Field of Study

Journal Information

Name

Page