Efficient spectral-temporal reconstruction of long-term satellite time series via temporal segments and mask-informed embedding

Jie Yang

doi:10.1016/j.rse.2026.115470

Efficient spectral-temporal reconstruction of long-term satellite time series via temporal segments and mask-informed embedding

2026-05-09

0

PRE

AI

J

Jie Yang

X

Xin Huang *

DOI：10.1016/j.rse.2026.115470

Original

Original request for help

Save

Abstract

En 中文

• MISTR enables efficient reconstruction of multi-decade time series. • Mask-informed embedding reduces missing-pattern bias in segments. • Channel-shared design supports flexible band configurations across sensors. • The proposed MISTR supports cross-sensor adaptation and multi-sensor synergy. • MISTR shows practical utility in classification, change detection, and phenology.

Keywords:

MISTR

temporal segments

mask-informed embedding

multi-sensor synergy

spectral-temporal reconstruction

Efficient spectral-temporal reconstruction of long-term satellite time series via temporal segments and mask-informed embedding

Abstract

Journal

Organization

Cited Papers

Citing Papers