Knightr0's Gambit is a partnership between the UCF student chapter of the Institute of Electrical
and Electronics Engineers (
UCF IEEE)
the UCF Artificial Intelligence club (
UCF AI)
to create an automatic chessboard (similar to
Harry Potter's Wizard Chess with less violence) powered by a custom chess AI. We plan to use computer vision to
detect the board state after each move (see
Computer Vision), build in natural language
processing capabilities to allow moving a piece without touching the board (see
Speech
Recognition), and we plan to build a web application to allow displaying the current game state as well as playing against
the chess board through a smartphone or computer interface (see
Web Application).
Chess and artificial intelligence have been closely related since the dawn of AI research
(
Turing). The state of the art for chess engines is constantly evolving,
but recent advancements include
AlphaZero, which uses reinforcement learning and self-play, and
Maia Chess, which uses supervised learning. We
aim to develop a custom chess AI inspired by AlphaZero and Maia that will incorporate machine learning and deep neural networks to
develop an understanding of the game of chess. For context, machine learning is a data-driven artificial intelligence methodology
and is generally divided into three categories: supervised, unsupervised (or self-supervised), and reinforcement learning. We plan
to primarily use reinforcement learning to develop our chess AI, but are also considering using supervised learning to improve the
AI by learning from high-level human chess games.
We are particularly inspired by 1) the Raspberry Turk project, which in turn was inspired by the famed
Mechanical Turk, a
chess automaton (later proven to be a hoax) that was built in the 1770s and 2) the
Phantom chess project, which in turn was inspired by the
Square Off chess board. There are many
examples of other similar projects, but we believe our approach is novel in combining and refining the following features:
In addition to the clear goal of pushing forward the state of the art for automatic chess boards and chess AI, a secondary
goal is one closely tied to the UCF IEEE and UCF AI mission statement: giving undergraduate students opportunities to apply concepts
from the classroom in a project-based environment.