KO | EN

GitLite

Minimalist Open Source Repository Search Engine

gitlite — search

#typescript #ai-agents #cli #claude-code #codex #developer-tools #dsh-plugin #deepseek-harness #windows #open-source #agent-skills #ai

MrPose

★ 33 Open GitHub ↗

An application to help with exercises for humans.

// repository documentation

Was this content helpful?

★ 0 (0 ratings)

# :weight_lifting_man: Mr.Pose :weight_lifting_woman:
[![License: MIT](https://img.shields.io/badge/License-MIT-yellow.svg)](https://github.com/Logeswaran123/MrPose/blob/main/LICENSE)
![alt text](https://github.com/Logeswaran123/MrPose/blob/main/data/images/Pose_DALL%C2%B7E.png "Pose DALL.E")

Hi human, 
I'm Mr.Pose! :robot: :wave: 
I help to estimate and measure exercises for humans. Lets do some exercise and keep the body fit. :muscle: :mechanical_arm:

## Description :scroll:

Exercises, in general, helps to maintain a physically fit body, loosen up joints, and relax muscles.

Mr.Pose is a visual analytics application that helps humans to track the accuracy of exercise, and count reps or to predict the exercise.

Mr.Pose can support 
* Pushup
* Plank
* Squat
* JumpingJack

For Pushup, Plank and JumpingJack, Mr.Pose will track each repetition provided that the exercise is accurately performed.
For Plank, Mr.Pose will track the time of plank position.

## General Requirements :mage_man:
1. Record a video of person doing a supported exercise.
2. Place the camera horizontally facing the person performing exercise (preferably at eye level).
3. Higher the resolution of video, higher the quality of estimation and measurement.

## Code Requirements :mage_woman:
```python
pip install -r requirements.txt
```

## How to run :running_man:
```python
python mrpose.py --video <path to video file> --exercise <exercise to be measured>
```
Note:

*<path to video file\>* - Path to the input video file with supported exercise 
*<exercise to be measured\>* - Exercise in input video file

**Optional Argument:** 
*--video* - Path to video source file. 
If argument is not provided, then Mr.Pose will launch webcam for live video. Currently, live webcam video works only for exercise prediction. 
*--exercise* - Choices are pushup, plank, squat, jumpingjack 
If argument is not provided, then Mr.Pose will **predict** the exercise done in the video. If argument is provided, then Mr.Pose will measure the exercise mentioned.

## Results :bar_chart:

:star: Pushups :star:

```python
python mrpose.py --video <path to video file> --exercise pushup
```

https://user-images.githubusercontent.com/36563521/188864300-bc43d096-c98f-48e4-b9cb-6bef6937f1ca.mp4

Input video source [here](https://www.pexels.com/video/woman-doing-push-ups-8472764/).
---

:star: Squats :star:

```python
python mrpose.py --video <path to video file> --exercise squat
```

https://user-images.githubusercontent.com/36563521/188866384-a19d3bb7-d6d2-47be-a71b-d27166dea395.mp4

Input video source [here](https://www.pexels.com/video/woman-exercising-while-wearing-a-face-mask-4265287/).
---

:star: Plank :star:

```python
python mrpose.py --video <path to video file> --exercise plank
```

https://user-images.githubusercontent.com/36563521/188879832-825cfd00-cfd3-4b9c-9d99-b73c4a06b8a9.mp4

Input video source [here](https://www.pexels.com/video/female-doing-planks-by-the-balcony-6152665/).
---

:star: Jumping Jacks :star:

```python
python mrpose.py --video <path to video file> --exercise jumpingjack
```

https://user-images.githubusercontent.com/36563521/188893014-afed7b58-6688-4e19-ba59-7a471c0fb7a0.mp4

Input video source [here](https://www.pexels.com/video/elderly-man-doing-jumping-jacks-outside-7299359/).
---

:star: Prediction :star:

```python
python mrpose.py --video <path to video file>
```

https://user-images.githubusercontent.com/36563521/190898949-a88b0d72-5a5a-43ec-b1dc-475a9744b6e6.mp4

---

## References :page_facing_up:

* [Mediapipe by Google](https://github.com/google/mediapipe).
* [GymLytics](https://github.com/akshaybahadur21/GymLytics).

Happy Learning! 😄

MrPose

Explore Similar Repositories

Was this content helpful?

Recent Feedback

Download README