Unsupervised Learning for Target Tracking and Background Subtraction in Satellite Imagery
This addresses the problem of high labeling costs for satellite imagery analysis, though it appears incremental as it builds on existing unsupervised techniques.
The paper tackles target tracking and background subtraction in satellite imagery using an unsupervised dual-model approach called Jekyll and Hyde, achieving competitive output quality with supervised methods without requiring labeled training data.
This paper describes an unsupervised machine learning methodology capable of target tracking and background suppression via a novel dual-model approach. ``Jekyll`` produces a video bit-mask describing an estimate of the locations of moving objects, and ``Hyde`` outputs a pseudo-background frame to subtract from the original input image sequence. These models were trained with a custom-modified version of Cross Entropy Loss. Simulated data were used to compare the performance of Jekyll and Hyde against a more traditional supervised Machine Learning approach. The results from these comparisons show that the unsupervised methods developed are competitive in output quality with supervised techniques, without the associated cost of acquiring labeled training data.