Journal
IEEE JOURNAL OF SELECTED TOPICS IN APPLIED EARTH OBSERVATIONS AND REMOTE SENSING
Volume 13, Issue -, Pages 6236-6248Publisher
IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC
DOI: 10.1109/JSTARS.2020.3030304
Keywords
Feature extraction; Hyperspectral imaging; Measurement; Legged locomotion; Task analysis; Associative learning; classification; domain adaptation; hyperspectral remote sensing
Categories
Funding
- National Natural Science Foundations of China [61771437, 61102104, 91442201]
Ask authors/readers for more resources
Associative learning-based domain adaptation approach is investigated for the classification of hyperspectral remote sensing images in this article. It employs the criterion of cycle consistency to achieve features that are both domain-invariant and discriminative. Two cross-domain similarity matrices based on network-generated features and probability predictions are introduced in the two-step transition procedure. The associative learning with feature and prediction-based similarity metrics is referred to as augmented associative learning (AAL). The AAL-based domain adaptation network does not require target labeled information and can achieve unsupervised classification of the target image. The experimental results using Hyperion and AVIRIS hyperspectral data demonstrated the efficiency of the proposed approach.
Authors
I am an author on this paper
Click your name to claim this paper and add it to your profile.
Reviews
Recommended
No Data Available