# MaskRCNN-Keypoint-Demo

**Repository Path**: kai_hui_tang/MaskRCNN-Keypoint-Demo

## Basic Information

- **Project Name**: MaskRCNN-Keypoint-Demo
- **Description**: Mask RCNN for the human pose estimation
- **Primary Language**: Unknown
- **License**: Not specified
- **Default Branch**: master
- **Homepage**: None
- **GVP Project**: No

## Statistics

- **Stars**: 2
- **Forks**: 0
- **Created**: 2019-11-09
- **Last Updated**: 2021-05-07

## Categories & Tags

**Categories**: Uncategorized

**Tags**: None

## README

# MaskRCNN-Keypoint-Demo
This repo is about Mask RCNN with human-keypoint. More info can be seen in this [repo](https://github.com/chrispolo/Keypoints-of-humanpose-with-Mask-R-CNN).

# Requirements
* Python 3.4+
* TensorFlow 1.3+
* keras 2.0.8+
* numpy, skimage, scipy, Pillow, cython, h5py
* cocoapi: `pip install "git+https://github.com/philferriere/cocoapi.git#egg=pycocotools&subdirectory=PythonAPI"`
* cv2: `pip install opencv-python`

# Run in notebook [![Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/FangYang970206/MaskRCNN-Keypoint-Demo/blob/master/demo.ipynb)
You can open the notebook in google colab, the environment can prepare in one cell and has free GPU.Just enjoy!

# Run in local
first, clone this repo,
```bash
$ git clone https://github.com/FangYang970206/MaskRCNN-Keypoint-Demo
```
then,
```bash
$ cd MaskRCNN-Keypoint-Demo
```
download the **pre-trained model**([baiduyun](https://pan.baidu.com/s/1C3o10VclGQy42UZu2dsl1w),[dropbox](https://www.dropbox.com/s/5ctrg3br94srrx9/mask_rcnn_coco.h5)) in the MaskRCNN-Keypoint-Demo folder.

finally, you have two choices:
```bash
$ python main.py --image path/to/image
```
or
```bash
$ python main.py --video path/to/video
```
Example:
```
$ python main.py --image media/ski.jpg
$ python main.py --video media/human.mp4
```
# Result
![demo.gif](gif/demo.gif)