drone Computer Vision Model
8550 Total Images
View All Images
Dataset Split
Train Set 88%
7485Images
Valid Set 8%
713Images
Test Set 4%
352Images
Preprocessing
Auto-Orient: Applied
Resize: Stretch to 224x224
Auto-Adjust Contrast: Using Histogram Equalization
Grayscale: Applied
Augmentations
Outputs per training example: 3
Flip: Horizontal
90° Rotate: Applied
Crop: 0% Minimum Zoom, 20% Maximum Zoom
Rotation: Between -15° and +15°
Shear: ±15° Horizontal, ±15° Vertical
Hue: Between -15° and +15°
Saturation: Between -25% and +25%
Brightness: Between -25% and +25%
Exposure: Between -25% and +25%
Blur: Up to 2px
Noise: Up to 5% of pixels
Cutout: 5 boxes with 10% size each
Mosaic: Applied