Unsupervised Learning for Target Tracking and Background Subtraction in Satellite Imagery

TL;DR

This paper introduces an unsupervised machine learning approach for target tracking and background subtraction in satellite imagery, using a dual-model system called Jekyll and Hyde, which performs competitively against supervised methods without needing labeled data.

Contribution

The paper presents a novel dual-model unsupervised framework for target tracking and background subtraction in satellite imagery, eliminating the need for labeled training data.

Findings

Unsupervised methods are competitive with supervised approaches in output quality.

The dual-model system effectively tracks targets and subtracts backgrounds.

The approach reduces the cost and effort of data labeling.

Abstract

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.