A Galaxy Image Augmentation Method Based on Few-shot Learning and Generative Adversarial Networks

作     者：Yiqi Yao Jinqu Zhang Ping Du Shuyu Dong 

作者机构：School of Artificial IntelligenceSouth China Normal UniversityFoshan 528225China School of Computer ScienceSouth China Normal UniversityGuangzhou 510631China Guangdong Construction Vocational Technology InstituteGuangzhou 511500China 

出 版 物：《Research in Astronomy and Astrophysics》 (天文和天体物理学研究（英文版）)

年 卷 期：2024年第24卷第3期

页      面：180-193页

核心收录：

学科分类：12[管理学] 1201[管理学-管理科学与工程(可授管理学、工学学位)] 07[理学] 081104[工学-模式识别与智能系统] 08[工学] 070401[理学-天体物理] 0835[工学-软件工程] 0704[理学-天文学] 0811[工学-控制科学与工程] 0812[工学-计算机科学与技术（可授工学、理学学位）] 

基　　金：supported by China Manned Space Program through its Space Application System,the National Natural Science Foundation of China(NSFC,grant Nos.11973022 and U1811464) the Natural Science Foundation of Guangdong Province(No.2020A1515010710) 

主　　题：techniques image processing-galaxies structure-galaxies general 

摘      要：Galaxy morphology classifications based on machine learning are a typical technique to handle enormous amounts of astronomical observation data,but the key challenge is how to provide enough training data for the machine learning *** this article proposes an image data augmentation method that combines few-shot learning and generative adversarial *** Galaxy10 DECaLs data set is selected for the experiments with consistency,variance,and augmentation effects being *** popular networks,including AlexNet,VGG,and ResNet,are used as examples to study the effectiveness of different augmentation methods on galaxy morphology *** results show that the proposed method can generate galaxy images and can be used for expanding the classification model’s training *** to comparative studies,the best enhancement effect on model performance is obtained by generating a data set that is 0.5–1 time larger than the original data ***,different augmentation strategies have considerably varied effects on different types of ***-GAN achieved the best classification performance on the ResNet network for In-between Round Smooth Galaxies and Unbarred Loose Spiral Galaxies,with F1 Scores of 89.54%and 63.18%,*** comparison reveals that various data augmentation techniques have varied effects on different categories of galaxy morphology and machine learning ***,the best augmentation strategies for each galaxy category are suggested.