Metrics constitute the foundation of Encord Active. In short, they serve the purpose of indexing data, labels, and model predictions based on various properties.
We have split the metrics into three main categories:
- Heuristic: operate on images or individual video frames and are heuristic in the sense that they mostly depend on the image content without labels.
- Geometric: operate on the geometries of objects like bounding boxes, polygons, and polylines.
- Semantic: operate with the semantic information of images or individual video frames.