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

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 partinformed 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.

Unsupervised Person Re-identification via Cross-Camera Similarity Exploration

“…Specifically, methods proposed to learn discriminative features from parts of pedestrian images achieve impressive performance [24,8,23]. For example, in [24], the feature maps are cut into uniform pieces for classification, and the partinformed features are assembled as the descriptor. A refined part pooling is further proposed to reinforce the within-part consistency in each part.…”

“…To assist the similarity measurement between the global feature, we propose to consider the similarity between part features (details) as well. Following [24], we extract the CNN feature map and divide it into p horizontal stripes. Each partition feature is then average pooled to be a part-level feature embedding.…”

“…Given a query image, person re-identification (re-ID) aims to match the person across multiple non-overlapped cameras. In the last few years, person re-ID has drawn increasing research attention [12,45,46,25,24,23], due to its wide range of applications such as finding people of interest (e.g., lost kids or criminals) and person tracking. However, most of the proposed methods are of supervised manner, which requires intensive manual labeling and is not applicable to real-world applications.…”

VMRFANet: View-specific Multi-Receptive Field Attention Network for Person Re-identification

“…Strip-based models: Recently, strip-based models have been proven to be effective in person re-ID. Part-based Convolutional Baseline (PCB) (Sun et al, 2018) equally slices the final feature map into horizontal strips. After refining part pooling, the extracted local features are jointly trained with classification losses and have been concatenated as the final feature.…”

“…• We investigate the idea of combining spatial-and channel-wise attention in a single module with various sized receptive filters, and then mount the module to a popular strip-based re-ID baseline (Sun et al, 2018) in a parallel way. We believe this is a more general form of attention module comparing to the ones in many existing structures that try to learn spatial-and channel-wise attention separately.…”