Geometric Cross-Modal Comparison of Heterogeneous Sensor Data

TL;DR

This paper introduces a geometric method for comparing heterogeneous sensor data streams, specifically aerial video and RF signals, using self-similarity matrices and advanced time-warping metrics to identify similar trajectories across modalities.

Contribution

The work presents a modality-agnostic technique for cross-modal comparison of sensor data streams using self-similarity matrices and a novel time-warping metric, demonstrating its effectiveness in trajectory classification.

Findings

Cross-modal SSM distances are smaller for same trajectory types.

The proposed method achieves high precision-recall in trajectory matching.

The approach is general and applicable to various modality combinations.

Abstract

In this work, we address the problem of cross-modal comparison of aerial data streams. A variety of simulated automobile trajectories are sensed using two different modalities: full-motion video, and radio-frequency (RF) signals received by detectors at various locations. The information represented by the two modalities is compared using self-similarity matrices (SSMs) corresponding to time-ordered point clouds in feature spaces of each of these data sources; we note that these feature spaces can be of entirely different scale and dimensionality. Several metrics for comparing SSMs are explored, including a cutting-edge time-warping technique that can simultaneously handle local time warping and partial matches, while also controlling for the change in geometry between feature spaces of the two modalities. We note that this technique is quite general, and does not depend on the choice of modalities. In this particular setting, we demonstrate that the cross-modal distance between SSMs corresponding to the same trajectory type is smaller than the cross-modal distance between SSMs corresponding to distinct trajectory types, and we formalize this observation via precision-recall metrics in experiments. Finally, we comment on promising implications of these ideas for future integration into multiple-hypothesis tracking systems.