Data Augmentation Based on Generative Adversarial Network with Mixed Attention Mechanism
Open Access
- 27 May 2022
- journal article
- research article
- Published by MDPI AG in Electronics
- Vol. 11 (11) , 1718
Abstract
Some downstream tasks often require enough data for training in deep learning, but it is formidable to acquire data in some particular fields. Generative Adversarial Network has been extensively used in data augmentation. However, it still has problems of unstable training and low quality of generated images. This paper proposed Data Augmentation Based on Generative Adversarial Network with Mixed Attention Mechanism (MA-GAN) to solve those problems. This method can generate consistent objects or scenes by correlating the remote features in the image, thus improving the ability to create details. Firstly, the channel-attention and the self-attention mechanism are added into the generator and discriminator. Then, the spectral normalization is introduced into the generator and discriminator so that the parameter matrix satisfies the Lipschitz constraint, thus improving the stability of the model training process. By qualitative and quantitative evaluations on small-scale benchmarks (CelebA, MNIST, and CIFAR-10), the experimental results show that the proposed method performs better than other methods. Compared with WGAN-GP (Improved Training of Wasserstein GANs) and SAGAN (Self-Attention Generative Adversarial Networks), the proposed method contributes to higher classification accuracy, indicating that this method can effectively augment the data of small samples.Keywords
This publication has 17 references indexed in Scilit:
- Interpretable Hyperspectral Artificial Intelligence: When nonconvex modeling meets hyperspectral remote sensingIEEE Geoscience and Remote Sensing Magazine, 2021
- Improved Techniques for Training Single-Image GANs2021
- Graph Convolutional Networks for Hyperspectral Image ClassificationIEEE Transactions on Geoscience and Remote Sensing, 2020
- PULSE: Self-Supervised Photo Upsampling via Latent Space Exploration of Generative Models2020
- SinGAN: Learning a Generative Model From a Single Natural Image2019
- Joint Discriminative and Generative Learning for Person Re-Identification2019
- A Style-Based Generator Architecture for Generative Adversarial Networks2019
- Deep learning in remote sensing scene classification: a data augmentation enhanced convolutional neural network frameworkGIScience & Remote Sensing, 2017
- An image enhancement technique combining sharpening and noise reductionIEEE Transactions on Instrumentation and Measurement, 2002
- Neocognitron: A self-organizing neural network model for a mechanism of pattern recognition unaffected by shift in positionBiological Cybernetics, 1980