2026-10-15T08:52:41Z https://www.scidb.cn/oai

10.57760/sciencedb.43091 2026-07-16T15:54:27Z

2026-07-16 Synthesir: Synthesizing infrared bands using the Wide-Field Imager and MSI (Multispectral Imager) optical sensors doi:10.57760/sciencedb.43091 en This dataset provides coregistered WFI–Sentinel 2 MSI patch pairs from AMAZONIA-1, CBERS-4, and CBERS-4A missions, enabling multi-satellite remote sensing applications such as spectral band synthesis and burned-area mapping. WFI patches contain Blue (1), Green (2), Red (3), and NIR (4) bands, while MSI patches include Red (1), Broad NIR (2), Narrow NIR (3), SWIR-1 (4), and SWIR-2 (5) bands. The Synthesir dataset includes temporally matched and near-temporal observations (0–7 days), along with continuous burned-area probability masks. A dedicated Sentinel-2 subset is also provided for training and evaluating deep learning models for burned-area semantic segmentation. By combining multi-resolution imagery, temporal variability, and pixel-level annotations, the dataset supports research on image-to-image translation, domain adaptation, weak-label learning, and deep learning-based Earth observation applications. CBERS-4; CBERS-4A; AMAZONIA-1; Sentinel-2; burned area mapping; deep learning Alisson Cleiton de Oliveira Rennan de Freitas Bezerra Marujo Thales Sehn Körting EMBARGO https://creativecommons.org/licenses/by/4.0/ dataset Science Data Bank