Aquarium
Search…
Stratified Metrics
Often you want to get a more nuanced understanding of how your model performs on specific slices of the labels. For example, in a self driving use case, you might want to understand how you perform on near vs far objects. Or you might have a fish-eye camera and what to understand how your model performs as objects approach the edge of the image.
With stratified metrics, you can filter the metrics view based on per-label attributes, looking at a specific 'stratum' of your label set. For example, if you attach an attribute called “range,” you can evaluate performance on just large or small labels independently.

Object Level Metadata Metrics Filtering

To enable label attributes metrics filtering, you can specify object metadata schema for the dataset or inference set using update_dataset_object_metadata_schema. For example,
import aquariumlearning as al
al_client = al.Client()
al_client.set_credentials(api_key=API_KEY)
schema_fields = [
{"name": "width_bucket", "type": "STRING"},
{"name": "width", "type": "FLOAT"}
]
al_client.update_dataset_object_metadata_schema(<PROJECT>, <DATASET>, schema_fields)
Then, use the "Object Level Filters" option to select the attribute and enter a value (ie medium) or numeric range (ie <10, <=10, =10,>=10, >10).

Example Usage (Deprecated)

stratified_metrics_definitions = [
al.StratifiedMetricsDefinition(
name='range_bucket',
ordered_values=['0-10', '10-20', '20-30', '30-40', '40-50', '50+']
)
]
...
range_bucket = get_cuboid_range_bucket((0,0,0), label['position'])
frame.add_label_3d_cuboid(user_attrs={'range_bucket': range_bucket}, **label)
...
range_bucket = get_cuboid_range_bucket((0,0,0), inference['position'])
inf_frame.add_inference_3d_cuboid(user_attrs={'range_bucket': range_bucket}, **inference)
...
al_client.create_project(
aquarium_project,
label_class_map,
stratified_metrics=stratified_metrics_definitions
)
Copy link
Outline
Object Level Metadata Metrics Filtering
Example Usage (Deprecated)