Edit this page

human-pose-estimation-3d-0001

Use Case and High-Level Description

Multi-person 3D human pose estimation model based on the Lightweight OpenPose and Single-Shot Multi-Person 3D Pose Estimation From Monocular RGB papers.

Example

_images/human-pose-estimation-3d-0001.jpg

Specification

Metric

Value

MPJPE (mm)

100.45

GFlops

18.998

MParams

5.074

Source framework

PyTorch*

Inputs

Image, name: data, shape: 1, 3, 256, 448 in the B, C, H, W format, where:

  • B - batch size

  • C - number of channels

  • H - image height

  • W - image width

Expected color order is BGR.

Outputs

The net outputs are three blobs:

  1. Name: features, shape: 1, 57, 32, 56 - coordinates in 3D space.

  2. Name: heatmaps, shape: 1, 19, 32, 56 - keypoint heatmaps.

  3. Name: pafs, shape: 1, 38, 32, 56 - keypoint pairwise relations (part affinity fields).

Download a Model and Convert it into Inference Engine Format

You can download models and if necessary convert them into Inference Engine format using the Model Downloader and other automation tools as shown in the examples below.

An example of using the Model Downloader:

python3 <omz_dir>/tools/downloader/downloader.py --name <model_name>

An example of using the Model Converter:

python3 <omz_dir>/tools/downloader/converter.py --name <model_name>
Prev Next