TRACER: Efficient Object Re-Identification in Networked Cameras through Adaptive Query Processing

TL;DR

Tracer is a novel VDBMS that improves object re-identification across camera networks by using adaptive query processing and a synthetic benchmark, significantly outperforming existing systems.

Contribution

It introduces an adaptive query processing framework with a recurrent network and probabilistic search model for efficient Re-ID in large camera networks.

Findings

Tracer outperforms state-of-the-art by 3.9x on average

It supports high recall constraints with dynamic sampling

A new synthetic benchmark for Re-ID datasets is proposed

Abstract

Efficiently re-identifying and tracking objects across a network of cameras is crucial for applications like traffic surveillance. Spatula is the state-of-the-art video database management system (VDBMS) for processing Re-ID queries. However, it suffers from two limitations. Its spatio-temporal filtering scheme has limited accuracy on large camera networks due to localized camera history. It is not suitable for critical video analytics applications that require high recall due to a lack of support for adaptive query processing. In this paper, we present Tracer, a novel VDBMS for efficiently processing Re-ID queries using an adaptive query processing framework. Tracer selects the optimal camera to process at each time step by training a recurrent network to model long-term historical correlations. To accelerate queries under a high recall constraint, Tracer incorporates a probabilistic adaptive search model that processes camera feeds in incremental search windows and dynamically updates the sampling probabilities using an exploration-exploitation strategy. To address the paucity of benchmarks for the Re-ID task due to privacy concerns, we present a novel synthetic benchmark for generating multi-camera Re-ID datasets based on real-world traffic distribution. Our evaluation shows that Tracer outperforms the state-of-the-art cross-camera analytics system by 3.9x on average across diverse datasets.