Dataset Versions
Versions
Test level 3 unnoise version
v12
· a year ago
test-Level 2
v11
· a year ago
Level-1
v10
· a year ago
Level-3
v9
· a year ago
Level-2
v8
· a year ago
2023-10-10 1:28pm
v7
· a year ago
2023-10-09 6:56pm
v6
· a year ago
2023-10-09 11:20am
v5
· a year ago
2023-10-08 7:59pm
v4
· a year ago
2023-10-08 6:38pm
v3
· a year ago
2023-10-08 6:35pm
v2
· a year ago
v2
2023-10-08 6:35pm
Generated on Oct 8, 2023
370 Total Images
View All Images
Dataset Split
Train Set 88%
324Images
Valid Set 8%
31Images
Test Set 4%
15Images
Preprocessing
Auto-Orient: Applied
Resize: Stretch to 480x480
Auto-Adjust Contrast: Using Histogram Equalization
Augmentations
Outputs per training example: 3
Flip: Horizontal, Vertical
90° Rotate: Clockwise, Counter-Clockwise, Upside Down
Crop: 0% Minimum Zoom, 25% Maximum Zoom
Rotation: Between -25° and +25°
Shear: ±20° Horizontal, ±25° Vertical
Grayscale: Apply to 25% of images
Hue: Between -100° and +100°
Saturation: Between -80% and +80%
Brightness: Between -40% and +40%
Exposure: Between -50% and +50%
Blur: Up to 2.5px
Noise: Up to 20% of pixels
Cutout: 3 boxes with 30% size each