JoJoPoseEstimation

Using OpenPose to recognize Poses from JoJo's Bizarre Adventure.

🎦 Video:

• https://youtu.be/NQvAatfP9nw

❔ What is it?

• Uses CMU's OpenPose to detect key body angles.
• To detect angles, I use the law of cosines (see prior GitHub repo that I made)
• Compares these frame angles and finds the closest JoJo pose.

⚡ Features:

• Detects body angles!
• Stores body angles!
• Compares input body angles to stored one!
• Returns the comparison!
• When returned, changes the input frame and plays music!
• Is a JoJo reference!

📦 Modules / Packages:

• numpy: https://numpy.org/devdocs/user/quickstart.html
• opencv: https://pypi.org/project/opencv-python/
• math: https://docs.python.org/3/library/math.html
• os: https://docs.python.org/3/library/os.html
• pygame: https://www.pygame.org/news
• time (in get_snapshot): https://docs.python.org/3/library/time.html

Basis:

• This idea is based on the work from: https://github.com/CMU-Perceptual-Computing-Lab/openpose

Guide-video:

• https://www.youtube.com/watch?v=09Tg1hXWGAM&ab_channel=log1

🔨 To do:

• Make it more accurate! Add more Jojo!