cst-tutorial-babybot

This repository provides a complete and reproducible Docker-based environment for training and integrating infant behavior models using the CST (Cognitive System Toolkit). It was developed as part of a tutorial presented at the ICDL (International Conference on Development and Learning).

🧠 Project Overview

This repository provides a complete and reproducible Docker-based environment for training and integrating infant behavior models using the CST (Cognitive System Toolkit). It was developed as part of a tutorial presented at the ICDL (International Conference on Development and Learning).

The environment supports:

Multimodal infant modeling with MIMo

Behavioral learning experiments using the BabyBench benchmark

Integration of cognitive models written in Java (CST) with simulation environments

A lightweight graphical interface accessible via web browser (noVNC)

✅ Requirements

This project runs entirely inside a Docker container, so the requirements are minimal on your host system.

---

🧱 Host System Requirements

These are required only on your local machine to build and run the Docker container.

Requirement	Recommended Version	Purpose
Docker	20.10+	To build and run the development environment
Operating System	Linux / macOS / Windows (with WSL2)	Linux or WSL2 is recommended for smoother support
Web Browser	Any (Chrome, Firefox, etc.)	To access the graphical interface via noVNC (port 8080)
Git	2.x	To clone BabyBench and MIMo repositories (optional)
CPU with AVX support	Yes	Recommended for PyTorch performance (MIMo models)

---

📦 Preinstalled Inside the Docker Container

Everything below is already included in the Dockerfile — no need to install manually:

🔧 Languages and Platforms

• Python 3.12 (via Miniconda)
• Java OpenJDK 17
• Gradle 7.6
• Maven
• CST-CLI (custom .deb package pre-installed)

🧠 Python Libraries

Installed via requirements.txt from BabyBench and MIMo:

• torch, numpy, matplotlib, pandas
• gymnasium, pygame, opencv-python, pyyaml
• websockify (for VNC proxy)
• stable-baselines3 (can be removed as per your Dockerfile)

🖥️ GUI and Developer Tools

• Xfce4 (lightweight desktop environment)
• x11vnc, xvfb, noVNC (browser-based GUI)
• Geany (code editor with syntax highlighting)
• Mousepad (lightweight text editor)
• xfce4-terminal (terminal emulator)

---

🧪 Additional Skills (Recommended for Usage)

• 🐍 Basic Python & Conda: To activate environments and run scripts
• ☕ Basic Java: To compile and run .java files
• 🧠 Interest in Developmental AI: To benefit from BabyBench simulations
• 🐳 Basic Docker knowledge: To build, run, and troubleshoot the container