Tsinghua university

sewer

Instance Segmentation

sewer Image Dataset

v1

2023-03-04 8:38pm

Generated on Mar 4, 2023

Dataset Split

Train Set 86%
462Images
Valid Set 9%
49Images
Test Set 5%
25Images

Preprocessing

Resize: Stretch to 640x640

Augmentations

Outputs per training example: 3
Flip: Horizontal, Vertical
Crop: 0% Minimum Zoom, 25% Maximum Zoom
Rotation: Between -45° and +45°
Noise: Up to 10% of pixels