person-reidentification-retail-0286

Use Case and High-Level Description

This is a person reidentification model for a general scenario. It uses a whole body image as an input and outputs an embedding vector to match a pair of images by the cosine distance. The model is based on the OSNet backbone with Linear Context Transform (LCT) blocks developed for fast inference. A single reidentification head from the 1/16 scale feature map outputs an embedding vector of 256 floats.

Example

_images/person-reidentification-retail-0286.jpg

Specification

Metric

Value

Market-1501 rank@1 accuracy

94.8 %

Market-1501 mAP

83.7 %

Pose coverage

Standing upright, parallel to image plane

Support of occluded pedestrians

YES

Occlusion coverage

<50%

GFlops

1.170

MParams

1.234

Source framework

PyTorch*

The cumulative matching curve (CMC) at rank-1 is accuracy denoting the possibility to locate at least one true positive in the top-1 rank. Mean Average Precision (mAP) is the mean across Average Precision (AP) of all queries. AP is defined as the area under the precision and recall curve.

Inputs

The net expects one input image of the shape 1, 3, 256, 128 in the B, C, H, W format, where:

  • B - batch size

  • C - number of channels

  • H - image height

  • W - image width

The expected color order is BGR.

Outputs

The net outputs a blob with the 1, 256 shape named reid_embedding which can be compared with other descriptors using the cosine distance.

Demo usage

The model can be used in the following demos provided by the Open Model Zoo to show its capabilities: