Dataset Versions

v2

2024-11-14 11:24pm

Generated on Nov 14, 2024

Popular Download Formats

Pascal VOC XML
Common XML annotation format for local data munging (pioneered by ImageNet).
PaliGemma
PaliGemma JSONL format used for fine-tuning PaliGemma, Google's open multimodal vision model.
CreateML JSON
CreateML JSON format is used with Apple's CreateML and Turi Create tools.
Other Formats
Choose another format.

Dataset Split

Train Set 88%
192Images
Valid Set 12%
27Images
Test Set %
0Images

Preprocessing

Auto-Orient: Applied
Static Crop: 14-76% Horizontal Region, 0-100% Vertical Region
Resize: Stretch to 640x640
Grayscale: Applied
Auto-Adjust Contrast: Using Histogram Equalization
Tile: 1 rows x 1 columns

Augmentations

Outputs per training example: 3
Flip: Horizontal, Vertical
90° Rotate: Clockwise, Counter-Clockwise, Upside Down
Crop: 0% Minimum Zoom, 30% Maximum Zoom
Rotation: Between -45° and +45°
Shear: ±15° Horizontal, ±15° Vertical
Grayscale: Apply to 22% of images
Hue: Between -24° and +24°
Saturation: Between -32% and +32%
Brightness: Between -22% and +22%
Exposure: Between -15% and +15%
Blur: Up to 0.7px
Noise: Up to 1.96% of pixels
Bounding Box: Flip: Horizontal, Vertical
Bounding Box: 90° Rotate: Clockwise, Counter-Clockwise, Upside Down
Bounding Box: Crop: 0% Minimum Zoom, 40% Maximum Zoom
Bounding Box: Rotation: Between -10° and +10°
Bounding Box: Shear: ±22° Horizontal, ±22° Vertical
Bounding Box: Brightness: Between -50% and +50%
Bounding Box: Exposure: Between -15% and +15%
Bounding Box: Blur: Up to 5px
Bounding Box: Noise: Up to 3.03% of pixels