Badr Moutalib

ScreenDetection

Instance Segmentation

ScreenDetection Image Dataset

v1

2023-03-31 3:14pm

Generated on Mar 31, 2023

Dataset Split

Train Set 87%
672Images
Valid Set 8%
64Images
Test Set 4%
33Images

Preprocessing

Auto-Orient: Applied
Resize: Stretch to 640x640

Augmentations

Outputs per training example: 3
Flip: Horizontal
90° Rotate: Clockwise, Counter-Clockwise
Crop: 0% Minimum Zoom, 45% Maximum Zoom
Rotation: Between -15° and +15°
Shear: ±15° Horizontal, ±15° Vertical
Hue: Between -7° and +7°
Saturation: Between -40% and +40%
Brightness: Between -57% and +57%
Exposure: Between -40% and +40%
Cutout: 12 boxes with 4% size each
Mosaic: Applied
Bounding Box: Flip: Horizontal
Bounding Box: Crop: 0% Minimum Zoom, 20% Maximum Zoom
Bounding Box: Brightness: Between -25% and +25%