What it does
Visual Object Detection identifies common objects in an image sent as a file upload or by URL. Send a file or a url, and choose which detector to use with requested_service (azure, google, or apyhub).
Use this service when you need to label images, filter uploads, or add image understanding to workflows that depend on object recognition. The file-based endpoint also accepts provider credentials where needed: Azure fields such as azure_key, azure_region, azure_endpoint, azure_account_id, and azure_access_token, plus google_key. The URL-based endpoint accepts nested azure and google credential objects, including Google service-account JSON and an optional google_language.
The response returns a data object with provider-specific results under azure, apyhub, and google. That lets you route results from the detector you selected without guessing at the response shape.
Typical uses include moderating user uploads, auto-tagging product photos, and building image search or inventory pipelines where you need basic object labels from a remote image or uploaded file.
| Provider (requested_service) | Atoms |
|---|
| Azure | 500 |
| Google | 500 |
| ApyHub | 2000 |