Data Augmentation in Emotion Classification Using Generative Adversarial Networks

Abstract

It is a difficult task to classify images with multiple class labels using only a small number of labeled examples, especially when the label (class) distribution is imbalanced. Emotion classification is such an example of imbalanced label distribution, because some classes of emotions like \emph{disgusted} are relatively rare comparing to other labels like {\it happy or sad}. In this paper, we propose a data augmentation method using generative adversarial networks (GAN). It can complement and complete the data manifold and find better margins between neighboring classes. Specifically, we design a framework with a CNN model as the classifier and a cycle-consistent adversarial networks (CycleGAN) as the generator. In order to avoid gradient vanishing problem, we employ the least-squared loss as adversarial loss. We also propose several evaluation methods on three benchmark datasets to validate GAN's performance. Empirical results show that we can obtain 5%~10% increase in the classification accuracy after employing the GAN-based data augmentation techniques.

References in corpus (11)

• Very Deep Convolutional Networks for Large-Scale Image Recognition
• SMOTE: Synthetic Minority Over-sampling Technique
• Conditional Generative Adversarial Nets
• Unsupervised Representation Learning with Deep Convolutional Generative Adversarial Networks
• Unpaired Image-to-Image Translation using Cycle-Consistent Adversarial Networks
• Deep Generative Image Models using a Laplacian Pyramid of Adversarial Networks
• Adversarially Learned Inference
• SeqGAN: Sequence Generative Adversarial Nets with Policy Gradient
• High-Performance Neural Networks for Visual Object Classification
• Least Squares Generative Adversarial Networks
• Box Drawings for Learning with Imbalanced Data

Cited by in corpus (15)

• AutoAugment: Learning Augmentation Policies from Data
• GAN Augmentation: Augmenting Training Data using Generative Adversarial Networks
• Multi-hop Federated Private Data Augmentation with Sample Compression
• Using generative adversarial networks to synthesize artificial financial datasets
• Online Hyper-parameter Learning for Auto-Augmentation Strategy
• Scale-aware Automatic Augmentation for Object Detection
• EfficientSeg: An Efficient Semantic Segmentation Network
• CLCC: Contrastive Learning for Color Constancy
• Learning Optimal Data Augmentation Policies via Bayesian Optimization for Image Classification Tasks
• Snore-GANs: Improving Automatic Snore Sound Classification with Synthesized Data
• Improving Augmentation and Evaluation Schemes for Semantic Image Synthesis
• Automatically Learning Data Augmentation Policies for Dialogue Tasks
• Enhancing Segment-Based Speech Emotion Recognition by Deep Self-Learning
• Data Augmentation with Variational Autoencoders and Manifold Sampling
• Object Recognition by a Minimally Pre-Trained System in the Process of Environment Exploration