Robot Agent Setup Guide

This project captures a drawing/photo, extracts a person-like skeleton from the image using a vision model API, and animates a dancing figure with Pygame.

1) Requirements

create and activate the virtual environment

conda create --name animated_drawings python=3.8.13
conda activate animated_drawings

# clone AnimatedDrawings and use pip to install
git clone https://github.com/facebookresearch/AnimatedDrawings.git
cd AnimatedDrawings
pip install -e .

2) Clone and open project

git clone https://github.com/Bobous2017/DrawBot-Studio/robot_agent.git
cd robot_agent

3) Create and activate a virtual environment

Windows (PowerShell)

python -m venv .venv
.\.venv\Scripts\Activate.ps1

If PowerShell blocks activation, run once:

Set-ExecutionPolicy -Scope CurrentUser RemoteSigned

macOS/Linux

python3 -m venv .venv
source .venv/bin/activate

4) Install dependencies

This project currently does not include a requirements.txt, so install directly:

pip install opencv-python numpy pygame requests python-dotenv openai

5) Configure environment variables

Create or edit .env in the project root:

Agent_API_Key=your_api_key_here
Agent_URL=https://ai/api/v1/chat/completions
Agent_MODEL=my_agent_module



Notes:
- `Agent_API_Key` and `Agent_URL` are required.
- `Agent_MODEL` is optional in code, but recommended to set explicitly.
- Do not commit real API keys to git.

## 6) Run the app

```bash
python main.py

Flow:

1. Camera window opens.
2. Press Space to capture image, or Esc to cancel.
3. Background is cut from the captured figure.
4. Vision model extracts joints.
5. Pygame opens and runs dance animation.

7) Quick environment check (optional)

python test_pygame.py

This prints installed library versions to verify setup.

Troubleshooting

• Camera cannot open:
  • Close other apps using webcam.
  • Check OS camera permissions.
• API request fails (400/401):
  • Verify Agent_API_Key, Agent_URL, and Agent_MODEL in .env.
• Pygame/OpenCV window does not show:
  • Make sure you are running in a desktop GUI session (not headless).
  • Update GPU/video drivers if needed.
• ModuleNotFoundError:
  • Confirm virtual environment is activated and dependencies are installed.

Project files

• main.py: Main app flow.
• capture_picture.py: Captures image from webcam.
• rigger_figure.py: Removes background and creates transparent figure PNG.
• vision_agent_module.py: Calls vision API to extract joints.
• dance_module.py: Animates skeleton/figure with Pygame.
• test_pygame.py: Basic dependency/version check.