Use Case
Ghegalearn
Ghegalearn’s output is a detection model that depends on your use case. The innovative way we use our training data pipeline is the key to the high detection and low false positive rates. Depending on your specific requirements, we create and adapt the training data we use until we have the most robust detection model.
Building the ground truth
Depending on the use case we can choose which format we want to use. A specialty is that we can change the annotation formats during the complete process on the fly.
Ghegapixel System
Panoptic Segmentation
Semantic Segmentation
3D Depth Map
YOLO Format
MS COCO Format
Proof of Concept
synthetic data for training
Train NN
inference
