加载中…
加载中…
This repository provides a Hugging Face–hosted version of a chess piece detection model originally created and trained by siromermer in the GitHub project:
Dynamic-Chess-Board-Piece-Extraction (GitHub link: https://github.com/siromermer/Dynamic-Chess-Board-Piece-Extraction).
The intention of this upload is to make the model easier to load in cloud environments such as Google Colab, especially for academic projects where instructors need to pull the weights directly without manual file handling. Full credit for the dataset preparation, initial training, and project concept belongs to the original author.
This model is based on YOLOv8 (Ultralytics), a state-of-the-art object detection architecture known for its efficiency and accuracy. The pretrained weights detect individual chess pieces on a board image.
The model was trained on labeled chessboard images using the class set defined by the original author.
The model predicts 12 chess piece categories, separated by color and type.
Class Names:
['black-bishop', 'black-king', 'black-knight', 'black-pawn', 'black-queen', 'black-rook',
'white-bishop', 'white-king', 'white-knight', 'white-pawn', 'white-queen', 'white-rook']
Class Dictionary:
{
0: 'black-bishop', 1: 'black-king', 2: 'black-knight', 3: 'black-pawn',
4: 'black-queen', 5: 'black-rook', 6: 'white-bishop', 7: 'white-king',
8: 'white-knight', 9: 'white-pawn', 10: 'white-queen', 11: 'white-rook'
}
from ultralytics import YOLO
# Load the model directly from Hugging Face
model = YOLO("https://huggingface.co/KanisornPutta/chess-model-yolov8m/resolve/main/chess-model-yolov8m.pt")
# Load a chessboard image
warped_image_path = "path/to/warped_board.jpg"
# Run inference
results = model(warped_image_path)
This work is not an original model but a re-hosting of pretrained weights created by:
This Hugging Face version exists solely for academic and educational convenience, with full acknowledgment of the model's creator.
Follow the license terms defined in the original repository. If none were specified, please contact the original author before using the model commercially.