Classifying Galaxy Morphologies with Few-Shot Learning

TL;DR

This paper demonstrates that few-shot learning with Siamese Networks outperforms traditional deep learning models in galaxy morphology classification, especially with limited training data, offering a more efficient and less biased approach.

Contribution

The study introduces the application of few-shot learning to galaxy morphology classification, showing significant accuracy improvements over conventional deep learning methods with fewer training images.

Findings

Few-shot learning achieves up to 21% higher accuracy than AlexNet with 1000 images.

Few-shot learning requires approximately 6300 images to reach 90% accuracy, less than ResNet_50's 13000 images.

Few-shot learning is suitable for classifying galaxy morphology and rare astrophysical objects with limited data.

Abstract

The taxonomy of galaxy morphology is critical in astrophysics as the morphological properties are powerful tracers of galaxy evolution. With the upcoming Large-scale Imaging Surveys, billions of galaxy images challenge astronomers to accomplish the classification task by applying traditional methods or human inspection. Consequently, machine learning, in particular supervised deep learning, has been widely employed to classify galaxy morphologies recently due to its exceptional automation, efficiency, and accuracy. However, supervised deep learning requires extensive training sets, which causes considerable workloads; also, the results are strongly dependent on the characteristics of training sets, which leads to biased outcomes potentially. In this study, we attempt Few-shot Learning to bypass the two issues. Our research adopts the dataset from Galaxy Zoo Challenge Project on Kaggle, and we divide it into five categories according to the corresponding truth table. By classifying the above dataset utilizing few-shot learning based on Siamese Networks and supervised deep learning based on AlexNet, VGG_16, and ResNet_50 trained with different volumes of training sets separately, we find that few-shot learning achieves the highest accuracy in most cases, and the most significant improvement is $21\%$ compared to AlexNet when the training sets contain 1000 images. In addition, to guarantee the accuracy is no less than 90\%, few-shot learning needs $\sim$6300 images for training, while ResNet_50 requires 13000 images. Considering the advantages stated above, foreseeably, few-shot learning is suitable for the taxonomy of galaxy morphology and even for identifying rare astrophysical objects, despite limited training sets consisting of observational data only.