Gallery Filter Network for Person Search

TL;DR

This paper introduces the Gallery Filter Network (GFN), a module that efficiently reduces gallery scenes in person search, combined with an improved SeqNeXt model, leading to significant performance improvements on standard datasets.

Contribution

The paper presents the GFN module for scene filtering and the SeqNeXt model, enhancing person search efficiency and accuracy over existing methods.

Findings

GFN effectively filters irrelevant gallery scenes across various scenarios.

SeqNeXt improves upon SeqNet with simpler architecture and better performance.

Combined SeqNeXt+GFN outperforms state-of-the-art on PRW and CUHK-SYSU datasets.

Abstract

In person search, we aim to localize a query person from one scene in other gallery scenes. The cost of this search operation is dependent on the number of gallery scenes, making it beneficial to reduce the pool of likely scenes. We describe and demonstrate the Gallery Filter Network (GFN), a novel module which can efficiently discard gallery scenes from the search process, and benefit scoring for persons detected in remaining scenes. We show that the GFN is robust under a range of different conditions by testing on different retrieval sets, including cross-camera, occluded, and low-resolution scenarios. In addition, we develop the base SeqNeXt person search model, which improves and simplifies the original SeqNet model. We show that the SeqNeXt+GFN combination yields significant performance gains over other state-of-the-art methods on the standard PRW and CUHK-SYSU person search datasets. To aid experimentation for this and other models, we provide standardized tooling for the data processing and evaluation pipeline typically used for person search research.