Beyond Part Models: Person Retrieval with Refined Part Pooling (and a Strong Convolutional Baseline)

TL;DR

This paper introduces a strong convolutional baseline for person retrieval called PCB, and a refined part pooling method RPP that improves part consistency and retrieval accuracy, surpassing previous state-of-the-art results.

Contribution

The paper proposes PCB as a competitive baseline and RPP for better part feature learning without external cues, advancing person retrieval methods.

Findings

PCB achieves state-of-the-art performance with a simple uniform partition.

RPP improves part consistency and boosts retrieval accuracy.

On Market-1501, the method surpasses previous best results in mAP and rank-1 accuracy.

Abstract

Employing part-level features for pedestrian image description offers fine-grained information and has been verified as beneficial for person retrieval in very recent literature. A prerequisite of part discovery is that each part should be well located. Instead of using external cues, e.g., pose estimation, to directly locate parts, this paper lays emphasis on the content consistency within each part. Specifically, we target at learning discriminative part-informed features for person retrieval and make two contributions. (i) A network named Part-based Convolutional Baseline (PCB). Given an image input, it outputs a convolutional descriptor consisting of several part-level features. With a uniform partition strategy, PCB achieves competitive results with the state-of-the-art methods, proving itself as a strong convolutional baseline for person retrieval. (ii) A refined part pooling (RPP) method. Uniform partition inevitably incurs outliers in each part, which are in fact more similar to other parts. RPP re-assigns these outliers to the parts they are closest to, resulting in refined parts with enhanced within-part consistency. Experiment confirms that RPP allows PCB to gain another round of performance boost. For instance, on the Market-1501 dataset, we achieve (77.4+4.2)% mAP and (92.3+1.5)% rank-1 accuracy, surpassing the state of the art by a large margin.