MediaPipe Pose Starter

A minimal Python starter project for MediaPipe Pose Landmark Detection.
Upload a video and get a processed video with the pose/skeleton overlaid.

---

1️⃣ Requirements

• Python 3.10 (Windows/Mac)
  ⚠️ MediaPipe does not fully support Python 3.12 yet.
• Virtual environment recommended
• pip installed

---

2️⃣ Project Structure

mediapipe-pose-starter/ │ ├── input_videos/ # Put your input videos here ├── output_videos/ # Processed videos will be saved here ├── src/ │ ├── main.py # Webcam demo │ └── process_video.py # Video processing script (run this) ├── .venv/ # Virtual environment (ignored by git) ├── requirements.txt └── .gitignore

---

3️⃣ Setup & Activate Environment

Windows (PowerShell)

# Create virtual environment (only once)
py -3.10 -m venv .venv

# Activate venv
.venv\Scripts\Activate.ps1

# Upgrade pip
pip install --upgrade pip

# Install dependencies
pip install -r requirements.txt

Mac

# Create virtual environment (only once)
python3.10 -m venv .venv

# Activate venv
source .venv/bin/activate

# Upgrade pip
pip install --upgrade pip

# Install dependencies
pip install -r requirements.txt

Always activate the virtual environment before running any scripts.
Check your Python version:

python --version
# should show Python 3.10.x

4️⃣ Run Video Processing

Put your input video in input_videos/ (e.g. sample.mp4).

Windows / Mac

Run python src/process_video.py

Processed video will be saved to output_videos/output.mp4

Console will show progress every 30 frames