Skip to main content

Working with Detectors

Explicitly create a new detector​

Typically you'll use the get_or_create_detector(name: str, query: str) method to find an existing detector you've already created with the same name, or create a new one if it doesn't exists. But if you'd like to force creating a new detector you can also use the create_detector(name: str, query: str) method

from groundlight import Groundlight

gl = Groundlight()

detector = gl.create_detector(name="your_detector_name", query="is there a hummingbird near the feeder?")

Retrieve an existing detector​

To work with a detector that you've previously created, you need to retrieve it using its unique identifier. This is typical in Groundlight applications where you want to continue to use a detector you've already created.

from groundlight import Groundlight

gl = Groundlight()

detector = gl.get_detector(id="your_detector_id")

Alternatively, you can retrieve a detector by its name:

from groundlight import Groundlight

gl = Groundlight()

detector = gl.get_detector_by_name(name="your_detector_name")

List your detectors​

To manage and interact with your detectors, you might need to list them. Groundlight provides a straightforward way to retrieve a list of detectors you've created. By default, the list is paginated to show 10 results per page, but you can customize this to suit your needs.

from groundlight import Groundlight

gl = Groundlight()

# Defaults to 10 results per page
detectors = gl.list_detectors()

# Pagination: 1st page of 5 results per page
detectors = gl.list_detectors(page=1, page_size=5)

[BETA] Create a Counting Detector​

So far, all of the detectors we've created have been binary classification detectors. But what if you want to count the number of objects in an image? You can create a counting detector to do just that. Counting detectors also return bounding boxes around the objects they count.

note

Counting Detectors are available on Pro, Business, and Enterprise plans.

from groundlight import ExperimentalApi

gl_experimental = ExperimentalApi()

detector = gl_experimental.create_counting_detector(name="your_detector_name", query="How many cars are in the parking lot?", max_count=20)

[BETA] Create a Multi-Class Detector​

If you want to classify images into multiple categories, you can create a multi-class detector.

from groundlight import ExperimentalApi

gl_experimental = ExperimentalApi()

class_names = ["Golden Retriever", "Labrador Retriever", "German Shepherd"]
detector = gl_experimental.create_multiclass_detector(
    name, query="What kind of dog is this?", class_names=class_names
)