OpenViSTAI · Agents · Blockchain · Crypto
Submit a project
Menu
AI Models

3Dmppe Posenet Release

3Dmppe Posenet Release is a Python and PyTorch implementation of the PoseNet component of a camera-distance-aware, top-down method for estimating 3D poses of multiple people from one RGB image. It includes demo, training, testing and visualization code.

Explore the practical guide ↓

Independent OpenViST project profile, based on linked sources.

3Dmppe Posenet Release — OpenViST project profile in AI Models.
OpenViST project illustration. See the linked sources for product details.
Understand the project

Practical overview

When it fits

Use it when you need research code for camera-distance-aware, top-down 3D multi-person pose estimation and can provide the required model, image and bounding-box and root-depth inputs.

Who it is for

  • Computer-vision researchers experimenting with multi-person 3D pose estimation.
  • Developers building or evaluating pose-estimation pipelines with supported datasets.

Documented uses

  • Estimate multiple people’s 3D poses from an RGB image using the provided PoseNet implementation.
  • Run a single-image demo and view 2D and 3D pose outputs.
  • Train or test the network and visualize pose-estimation results.
  • Work with listed pose datasets or prepare a custom dataset in MS COCO format.
Before you choose

Requirements & limits

The public GitHub repository lists an MIT license.

  • This repository provides the PoseNet part of the research method, not a hosted pose-estimation service.
  • The quick demo requires pretrained weights and manually configured bounding-box and root-depth lists.
  • The README lists PyTorch, CUDA, cuDNN, Anaconda and the COCO API as dependencies.
  • Custom datasets must be converted to MS COCO format.

Check current requirements in the official documentation. Source-based guidance does not mean OpenViST has run or independently tested this project.

Source-based practical guide

Run the documented single-image demo

Use the repository’s demo to generate 2D and 3D pose visualizations for an image.

Before you start

  • PyTorch, CUDA, cuDNN, Anaconda and the COCO API.
  • A pretrained PoseNet snapshot and an input image.
  • Bounding-box and root-depth lists for the demo.

Steps

  1. Place the input image and pretrained snapshot in the demo folder.

  2. Set bbox_list and root_depth_list in demo.py.

  3. Run the command below to use GPU 0.

    Documented command
    python demo.py --gpu 0 --test_epoch 24

Expected outcome

The README says the demo produces output_pose_2d.jpg and opens a window showing the 3D pose.

Guide limitations

  • The demo requires bounding-box and root-depth values to be set.
  • The documented development environment used Python 3.6.5, Ubuntu 16.04, CUDA 9.0, cuDNN 7.1 and two NVIDIA 1080Ti GPUs; this is not a compatibility guarantee.
Documentation 1 ↗
Evidence and transparency

Sources & review

Official repository ↗
Source record details
Source availability
Active
Sources checked
15 October 2026
Original website title
mks0601/3DMPPE_POSENET_RELEASE · Official homepage title · retained in its source language
How we review sourcesSuggest a correction