Automatic Decision Shaping

Automatic decision shaping (simulation 1). A system of thermal printers voting to decide which pattern to print.

Agent Voting Simulation for Thermal Printer Pattern Selection

This program simulates agents voting collectively to decide which pattern the associated thermal printers will print. Each agent is equipped with a neural network model to make decisions based on given image candidates. The winning pattern is then printed by all the thermal printers.

Requirements

• Python 3.x
• TensorFlow
• OpenCV (cv2)

Installation

1. Clone or download the repository.
2. Install the required Python packages by running:

   pip install -r requirements.txt
   

Usage

Run the program with the following command:

 python3 main.py [nb_agents_to_setup] [delay_between_each_vote_in_ms] [training_nb_files] [training_nb_epoch] [--http] [--reset]

• nb_agents_to_setup: Number of agents participating in the simulation.
• delay_between_each_vote_in_ms: Delay between each vote in milliseconds.
• training_nb_files: Number of training files for neural network training.
• training_nb_epoch: Number of epochs for neural network training.
• --http: (Optional) Enable HTTP output mode for visualization (http://localhost:8000).
• --reset: (Optional) Reset and retrain neural network models.

Example:

  python3 main.py 5 10000 1000 10 --http

Features

• Neural Network Training: Trains neural network models based on self generated training data.
• Agent Voting: Simulates agent decision-making based on models trained with preferences.
• Thermal Printer Integration: Print the winning pattern for every agent, using the associated thermal printer.
• HTTP Visualization: Optional mode to visualize a simulation results via HTTP.