MatteoM95 / Multitask-segmentation-from-satellite-imagery-for-burned-area-delineation-and-severity-estimation

A multi-task learning semantic segmentation approach is employed for targeting both wildfire delineation and burn severity estimation. By exploiting a large dataset of images from past wildfires and integrating both tasks into a single network can exceed the state-of-the-art results on this topic.

Repository from Github https://github.comMatteoM95/Multitask-segmentation-from-satellite-imagery-for-burned-area-delineation-and-severity-estimationRepository from Github https://github.comMatteoM95/Multitask-segmentation-from-satellite-imagery-for-burned-area-delineation-and-severity-estimation

Multitask-segmentation-from-satellite-imagery-for-burned-area-delineation-and-severity-estimation

Multitask segmentation from satellite imagery for burned area delineation and severity estimation

About

A multi-task learning semantic segmentation approach is employed for targeting both wildfire delineation and burn severity estimation. By exploiting a large dataset of images from past wildfires and integrating both tasks into a single network can exceed the state-of-the-art results on this topic.

License:Apache License 2.0


Languages

Language:Jupyter Notebook 91.0%Language:Python 9.0%Language:Shell 0.0%