Dual-window transformer framework with pyramid structure and constrained self-attention for hyperspectral anomaly detection | Journal of the European Optical Society-Rapid Publications

Open Access

Table 2

AUC under different dual-window.

(HEADS = 1, DEPTH = 5)
Dataset	(win_in, win_out)
Dataset	(1, 3)	(1, 5)	(1, 7)	(3, 5)	(3, 7)	(5, 7)	Max–Min
Holly	0.9998	0.9997	0.9995	0.9997	0.9996	0.9997	0.0003
Cement Street	0.9392	0.9490	0.9464	0.9454	0.9429	0.9394	0.0098
Jungle-I	0.9807	0.9786	0.9788	0.9793	0.9789	0.9791	0.0021
Jungle-II	0.9452	0.9389	0.9407	0.9455	0.9383	0.9469	0.0086
Average	0.9661	0.9666	0.9664	0.9675	0.9649	0.9663	0.0026

Current usage metrics show cumulative count of Article Views (full-text article views including HTML views, PDF and ePub downloads, according to the available data) and Abstracts Views on Vision4Press platform.

Data correspond to usage on the plateform after 2015. The current usage metrics is available 48-96 hours after online publication and is updated daily on week days.

Initial download of the metrics may take a while.