MovAl

KO | EN

gitlite — search

#python #java #python3 #arduino #golang #machine-learning #rust #html #flask #javascript #seismology #nodejs

★ 20 Open GitHub ↗

Integrated pipeline of YOLO with instance segmentaion and contour for multiple animal pose estimation

// repository documentation

Was this content helpful?

★ 0 (0 ratings)

# MovAl (Move Altogether)
![background_image](https://github.com/user-attachments/assets/d3af7702-ae83-4f63-95d7-0907b87eeac7)

**MovAl:** Move Altogether is an integrated pipeline of YOLO (pose) and Cutie (Instance segmentation) for multi animal key point detection. 
This pipeline overcomes the Id-switching problem of existing multi-animal tracking methods and give better key point detection quality. Try it for your multi-instance!

![그림2](https://github.com/user-attachments/assets/c5652ecb-3ee8-402a-8a71-4e1059db3ea8)
![그림3](https://github.com/user-attachments/assets/c4d4b78c-3bfe-4d05-8835-f48ba638381d)

[🔗 See more tracking results!](https://github.com/coldlabkaist/MovAl/blob/main/tutorial/Tracking_Result.md)

## Requirement
- CUDA 11.8 (or 12.1, 12.8)
- Supports Windows 11, Ubuntu 22.04 environment.

### Installation Tutorial
[0. Installation](https://github.com/coldlabkaist/MovAl/blob/main/tutorial/0_Installation.md)

### Tutorial on using the MovAl
[1. Create Project](https://github.com/coldlabkaist/MovAl/blob/main/tutorial/1_Create_Project.md)

[2. Preprocess](https://github.com/coldlabkaist/MovAl/blob/main/tutorial/2_Preprocess.md)

[3. Labelary](https://github.com/coldlabkaist/MovAl/blob/main/tutorial/3_Labelary.md)

[4. Pose Estimation](https://github.com/coldlabkaist/MovAl/blob/main/tutorial/4_Pose_Estimation.md)

## News
[Update Notes!](https://github.com/coldlabkaist/MovAl/blob/main/CHANGELOG.md )

If your local MovAl version is outdated compared to the latest release version on github, you can see a message recommending an update when running the code. You can simply update it using the following command
```
cd (MovAl_folder)
python update_moval.py
# or
python update_moval.py v1.1.1
```
If no version is provided, MovAl updates to the latest release tag. If a version is provided, MovAl checks out that tag.

## Liscense
MovAl basically follow the MIT license. We allows free use for academic/research purposes, but not for commercial purposes.
We recommends using Cutie and YOLO as part of our pipeline. The usage rights for each part follow the license of each program. 
- Cutie : https://github.com/hkchengrex/Cutie
- YOLO : https://www.ultralytics.com/

MovAl

Explore Similar Repositories

Was this content helpful?

Recent Feedback

Download README