Spatio-Visual Fusion-Based Person Re-Identification for Overhead Fisheye Images
This addresses the problem of person re-identification in crowded indoor spaces using overhead fisheye cameras, which is incremental as it builds on existing methods for a specific domain.
The paper tackles person re-identification for overhead fisheye cameras, a less explored scenario, by proposing a multi-feature framework that combines deep-learning, color-based, and location-based features, achieving an 18% improvement over appearance-based deep-learning methods and a 3% improvement over location-based methods in matching accuracy.
Person re-identification (PRID) has been thoroughly researched in typical surveillance scenarios where various scenes are monitored by side-mounted, rectilinear-lens cameras. To date, few methods have been proposed for fisheye cameras mounted overhead and their performance is lacking. In order to close this performance gap, we propose a multi-feature framework for fisheye PRID where we combine deep-learning, color-based and location-based features by means of novel feature fusion. We evaluate the performance of our framework for various feature combinations on FRIDA, a public fisheye PRID dataset. The results demonstrate that our multi-feature approach outperforms recent appearance-based deep-learning methods by almost 18% points and location-based methods by almost 3% points in matching accuracy. We also demonstrate the potential application of the proposed PRID framework to people counting in large, crowded indoor spaces.