metadata
license: apache-2.0
tags:
- chess
- segmentation
- unet
- computer-vision
pipeline_tag: image-segmentation
Chess Board Segmentation V6
Model Description
Breakthrough V6 UNet model for precise chess board segmentation and boundary detection.
Performance
- Dice Score: 0.9391
- Input Size: 256x256
- Architecture: Custom UNet V6 with 32 base channels
Usage
import torch
from transformers import AutoModel
model = AutoModel.from_pretrained("yamero999/chess-board-segmentation-v6")
# Process 256x256 RGB images
Training
Trained on diverse chess board images with various angles, lighting conditions, and board styles.