In order to join virtual sessions, you must be registered and logged-in(Were you registered for the in-person meeting in Denver? if yes, just log in.) 
Note: All session times are in Mountain Daylight Time.

A Review of Deep Learning in Remote Sensing: Achievements and limitations

Authors: Zhijie Zhang*, UCONN, Chuanrong Zhang, UCONN, Weidong Li, UCONN
Topics: Remote Sensing, Geographic Information Science and Systems
Keywords: Remote Sensing, Deep Learning, Constitutional Neural Network, Review
Session Type: Poster
Presentation File: No File Uploaded


Traditional Machine Learning methods have changed the landscape of the field of Remote Sensing (RS), as a currently trending sub-field of Machine Learning, Deep Learning (DL) continues to provide more accurate and efficient methods for RS data processing. Due to the development of Convolutional Neural Network (CNN) and its strong ability to process images, many RS scientists had many successful attempts to apply CNN into RS imagery processing. The objective of this paper is to review and discuss how CNN based DL methods are applied and fine-tuned in different areas of RS. First, we divide the application area of RS by data type that was used in the study (e.g. Natural Image, Multi-spectral, Hyperspectral) and briefly introduce how those data were processed before DL. Then we further discuss what kind of application was that type of data used in and how are they combined with DL. Finally, we summarize and discuss current issues and challenges of applying DL methods into RS. We hope that through this review, RS researchers can have a clearer picture of the development of DL in the field of RS and thus further assist their future research.

To access contact information login