Pawnimation

Pawnimation – Cinematic Chess Visualization in Unreal Engine 5

Pawnimation is a cinematic visualization system that transforms chess matches into immersive 3D experiences.
Built in Unreal Engine 5, it loads games from PGN files and recreates them move by move, with animations, sound effects, and dynamic camera movements that make each turn feel like a scene from a movie.

Developed as part of a Master’s Degree thesis in Computer Engineering, the project explores how cinematic techniques and reinforcement learning can enhance the visualization of abstract games like chess.

Main Features

Cinematic Replay – Recreates complete chess matches with smooth animations and visual effects.
AI Camera Control – A Reinforcement Learning agent dynamically chooses camera positions to achieve aesthetic, film-like framing.
Destruction Effects – Captured pieces are replaced with physics-based geometry collections that shatter on impact.
Stylized Environment – The chessboard floats in a vast cosmic background, enhancing the epic and surreal tone of the match.
Immersive Soundtrack – Original sound design to match the intensity of the moves.

Technical Highlights

• Engine: Unreal Engine 5.5

• Implementation: Blueprint

• Experimental Component: Reinforcement Learning agent (PPO-based) trained to find cinematic camera angles

• PGN File Support: Loads real chess games

• Compatibility: Windows and macOS

About the Project

This work was created as part of my Master’s Thesis:
“Pawnimation – Cinematic Chess Visualization in Unreal Engine 5 with Reinforcement Learning Camera Control”

The goal was to explore how AI-driven cinematography can bring expressive storytelling into algorithmic visualization — blending logic, art, and technology.

Downloads

Full Project (20 GB): Download on MEGA

GitHub Repository: View on GitHub

Author

Simone Di Cocco
Master’s Degree in Engineering in Computer Science (2025)
Sapienza Università di Roma