Handling unexpected inputs: incorporating source-wise out-of-distribution detection into SAR-optical data fusion for scene classification
The fusion of synthetic aperture radar (SAR) and optical satellite data is widely used for deep learning based scene classification. Counter-intuitively such neural networks are still sensitive to changes in single data sources, which can lead to unexpected behavior and a significant drop in perform...
Main Authors: | Gawlikowski, J. (Author), Niebling, J. (Author), Saha, S. (Author), Zhu, X.X (Author) |
---|---|
Format: | Article |
Language: | English |
Published: |
Springer Science and Business Media Deutschland GmbH
2023
|
Subjects: | |
Online Access: | View Fulltext in Publisher View in Scopus |
Similar Items
-
Software modification and implementation for, and analysis of, lidar data
by: Apte, Madhav Vasudeo, 1958-
Published: (1988) -
Meta-Analysis of Wetland Classification Using Remote Sensing: A Systematic Review of a 40-Year Trend in North America
by: Masoud Mahdianpari, et al.
Published: (2020-06-01) -
Multisensor remote sensing data integration
by: Nedim Kulo
Published: (2019-12-01) -
Detecting Ships in the New Zealand Exclusive Economic Zone: Requirements for a Dedicated SmallSat SAR Mission
by: Jan Krecke, et al.
Published: (2021-01-01) -
Identification of layered cloud occurrences from the Lidar In-Space Technology Experiment and advanced very high resolution radiometer imagery
by: Stevermer, Amy J.
Published: (2012)