segment-selector

A naive polygon segment selector using OpenAI Agents SDK function calling for prompt-based segment selection. The repository explores possibilities for an AI-assisted geometric task.

Selected Segments by Agent

Installation

This repository uses the image named python:3.12 for running devcontainer.

1. Ensure you have Docker and Visual Studio Code with the Remote - Containers extension installed.

2. Clone the repository.

   git clone https://github.com/PARKCHEOLHEE-lab/segment-selector.git
   

3. Open the project with VSCode.

4. When prompted at the bottom left on the VSCode, click Reopen in Container or use the command palette (F1) and select Remote-Containers: Reopen in Container.

5. VS Code will build the Docker container and set up the environment.

6. Once the container is built and running, you're ready to start working with the project.

File Details

src

• selector.py: Implementation of the segment selector agent
• plotter.py: Visualizer for selected segments
• testsets.py: Test cases with polygon boundaries and selection prompts
• main.py: Entry point for running the segment selector

runs

• Testcase*.png: Visualization results of different test cases

tests

• test_selection.py: Unit tests about whether the agent's decisions are intended results

How to run

1. Set up OpenAI API key:

   export OPENAI_API_KEY=your_api_key
   

2. Run the main script:

   python segment_selector/src/main.py
   

3. Adding New Tools:

   • Use @function_tool decorator to add new tools for the agent
   • Define static methods in SegmentSelector class to automatically register tools

4. Configuration:

   • Adjust hyperparameters in the SegmentSelectorConfiguration class
   • Modify MODEL_SETTINGS in the SegmentSelectorConfiguration for different behaviors