Escuela Superior politecnica del chimborazo

GastricX

Classification

GastricX Image Dataset

v1

2023-12-24 5:49pm

Generated on Dec 24, 2023

Dataset Split

Train Set 88%
2070Images
Valid Set 8%
197Images
Test Set 4%
98Images

Preprocessing

Auto-Orient: Applied
Resize: Stretch to 640x640

Augmentations

Outputs per training example: 3
Flip: Horizontal, Vertical
90° Rotate: Clockwise, Counter-Clockwise, Upside Down
Crop: 0% Minimum Zoom, 10% Maximum Zoom
Rotation: Between -16° and +16°
Shear: ±28° Horizontal, ±0° Vertical
Grayscale: Apply to 58% of images
Hue: Between -36° and +36°
Saturation: Between -74% and +74%
Brightness: Between -42% and +42%
Exposure: Between -30% and +30%
Blur: Up to 4px
Noise: Up to 14% of pixels
Cutout: 10 boxes with 8% size each