DMARTs Computer Vision Dataset
96 Total Images
View All Images
Dataset Split
Train Set 97%
93Images
Valid Set 2%
2Images
Test Set 1%
1Images
Preprocessing
Auto-Orient: Applied
Resize: Fit within 640x640
Augmentations
Outputs per training example: 3
Noise: Up to 1.81% of pixels
Bounding Box: Shear: ±10° Horizontal, ±10° Vertical