Disentangling the Benefits of Self-Supervised Learning to Deployment-Driven Downstream Tasks of Satellite Images (Student Abstract)

Authors

  • Zhuo Deng Tsinghua Shenzhen International Graduate School, Shenzhen, Guangdong, China Ping An Technology, Shenzhen, Guangdong, China
  • Yibing Wei University of Wisconsin-Madison, Madison, WI, USA PAII Inc., Palo Alto, CA, USA
  • Mingye Zhu University of Science and Technology of China, Hefei, Anhui, China Ping An Technology, Shenzhen, Guangdong, China
  • Xueliang Wang Tsinghua Shenzhen International Graduate School, Shenzhen, Guangdong, China
  • Junchi Zhou Tsinghua Shenzhen International Graduate School, Shenzhen, Guangdong, China
  • Zhicheng Yang PAII Inc., Palo Alto, CA, USA
  • Hang Zhou PAII Inc., Palo Alto, CA, USA
  • Zhenjie Cao Ping An Technology, Shenzhen, Guangdong, China Tsinghua Shenzhen International Graduate School, Shenzhen, Guangdong, China
  • Lan Ma Tsinghua Shenzhen International Graduate School, Shenzhen, Guangdong, China
  • Mei Han PAII Inc., Palo Alto, CA, USA
  • Jui-Hsin Lai PAII Inc., Palo Alto, CA, USA

DOI:

https://doi.org/10.1609/aaai.v37i13.26959

Keywords:

Self-supervised Learning, Satellite Image, Deployment-driven

Abstract

In this paper, we investigate the benefits of self-supervised learning (SSL) to downstream tasks of satellite images. Unlike common student academic projects, this work focuses on the advantages of the SSL for deployment-driven tasks which have specific scenarios with low or high-spatial resolution images. Our preliminary experiments demonstrate the robust benefits of the SSL trained by medium-resolution (10m) images to both low-resolution (100m) scene classification case (4.25%↑) and very high-resolution (5cm) aerial image segmentation case (1.96%↑), respectively.
AAAI-23 Proceedings Cover

Downloads

Published

2024-07-15

How to Cite

Deng, Z., Wei, Y., Zhu, M., Wang, X., Zhou, J., Yang, Z., Zhou, H., Cao, Z., Ma, L., Han, M., & Lai, J.-H. (2024). Disentangling the Benefits of Self-Supervised Learning to Deployment-Driven Downstream Tasks of Satellite Images (Student Abstract). Proceedings of the AAAI Conference on Artificial Intelligence, 37(13), 16198-16199. https://doi.org/10.1609/aaai.v37i13.26959

Issue

Vol. 37 No. 13: AAAI-23 Special Programs, IAAI-23, EAAI-23, Student Papers and Demonstrations

Section

AAAI Student Abstract and Poster Program