Dataset Versions

v1

is_pipe1764

Generated on May 19, 2023

Dataset Split

Train Set 92%
4233Images
Valid Set 5%
246Images
Test Set 2%
107Images

Preprocessing

Auto-Orient: Applied

Augmentations

Outputs per training example: 3
Flip: Horizontal, Vertical
Rotation: Between -10° and +10°
Shear: ±15° Horizontal, ±15° Vertical
Saturation: Between -24% and +24%
Exposure: Between -25% and +25%
Bounding Box: Exposure: Between -19% and +19%
Bounding Box: Blur: Up to 2.5px
Bounding Box: Noise: Up to 5% of pixels