Multimodal Representation Learning and Set Attention for LWIR In-Scene Atmospheric Compensation
A multimodal generative modeling approach combined with permutation-invariant set attention is investigated in this article to support long-wave infrared (LWIR) in-scene atmospheric compensation. The generative model can produce realistic atmospheric state vectors (T, H<sub>2</sub>O, O&l...
Main Authors: | , , , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
IEEE
2021-01-01
|
Series: | IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing |
Subjects: | |
Online Access: | https://ieeexplore.ieee.org/document/9242299/ |