Unsupervised Change Detection of Extreme Events Using ML On-Board
This addresses the bottleneck of downlink constraints in satellite applications like disaster management, offering a practical solution for faster data availability.
The paper tackles the problem of slow satellite data analysis for disaster management by introducing RaVAEn, an unsupervised change detection method that processes data on-board to prioritize downlink, reducing response time and outperforming pixel-wise baselines.
In this paper, we introduce RaVAEn, a lightweight, unsupervised approach for change detection in satellite data based on Variational Auto-Encoders (VAEs) with the specific purpose of on-board deployment. Applications such as disaster management enormously benefit from the rapid availability of satellite observations. Traditionally, data analysis is performed on the ground after all data is transferred - downlinked - to a ground station. Constraint on the downlink capabilities therefore affects any downstream application. In contrast, RaVAEn pre-processes the sampled data directly on the satellite and flags changed areas to prioritise for downlink, shortening the response time. We verified the efficacy of our system on a dataset composed of time series of catastrophic events - which we plan to release alongside this publication - demonstrating that RaVAEn outperforms pixel-wise baselines. Finally we tested our approach on resource-limited hardware for assessing computational and memory limitations.