https://github.com/wonder-tree/posecamera-cpp

https://github.com/wonder-tree/posecamera-cpp

Last synced: 8 months ago
JSON representation

• Host: GitHub
• URL: https://github.com/wonder-tree/posecamera-cpp
• Owner: Wonder-Tree
• License: apache-2.0
• Created: 2021-03-24T07:11:25.000Z (over 5 years ago)
• Default Branch: main
• Last Pushed: 2021-06-07T07:03:08.000Z (about 5 years ago)
• Last Synced: 2025-01-19T08:18:31.309Z (over 1 year ago)
• Language: C++
• Size: 16.9 MB
• Stars: 0
• Watchers: 3
• Forks: 1
• Open Issues: 0
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

          
[![CodeFactor](https://www.codefactor.io/repository/github/wonder-tree/posecamera-cpp/badge)](https://www.codefactor.io/repository/github/wonder-tree/posecamera-cpp)

# PoseCamera(cpp)

## Realtime Humanpose Estimation in C++

This is a realtime humanpose estimation using libtorch. This project can runs realtime (16fps) on CPU and about 50fps on GPU. 

## Install

> use `trace_model.py` file to trace your pytorch model for libtorch c++

In order to run this demo, you have to follow these steps:

1. Install thor which provide visualization utils.

   ```

   # Download libtorch

   # install thor

   git clone http://github.com/jinfagang/thor

   cd thor

   mkdir build && cd build

   cmake ..

   make -j8

   sudo make install

   ```

   

2. make executable file

   ```

   mkdir build

   cd build

   cmake .. && make -j4

   ```

      

And that is all!  **make sure you have copied libtorch to your home folder**.