jintanaboonp

MEDBOX

Classification

MEDBOX Image Dataset

v1

2024-06-10 2:47pm

Generated on Jun 10, 2024

Dataset Split

Train Set 85%
330Images
Valid Set 10%
37Images
Test Set 5%
20Images

Preprocessing

Auto-Orient: Applied
Resize: Stretch to 640x640
Auto-Adjust Contrast: Using Contrast Stretching

Augmentations

Outputs per training example: 3
90° Rotate: Clockwise, Counter-Clockwise, Upside Down
Rotation: Between -15° and +15°
Saturation: Between -42% and +42%
Brightness: Between -20% and +20%
Exposure: Between -23% and +23%
Noise: Up to 0.69% of pixels