Pretrained Image Analysis Models

Vision provides pretrained image analysis AI models that let you to find and tag objects, text, and entire scenes in images.

Pretrained models let you use AI with no data science experience. Provide an image to the Vision service and get back information about the objects, text, scenes, and any faces in the image without having to create your own model.

Use Cases

Here are several use cases for pretrained image analysis models.

Digital asset management
Tag digital media-like images for better indexing and retrieval.
Scene monitoring
Detect if items are on retail shelves, vegetation is growing in the surveillance image of a power line, or if trucks are available at a lot for delivery or shipment.
Face detection
  • Privacy: Hide identities by adding a blur to the image using face location information returned through the face detection feature.
  • Prerequisite for Biometrics: Use the facial quality score to decide if a face is clear and unobstructed.
  • Digital asset management: Tag images with facial information for better indexing and retrieval.

Supported Formats

Vision supports several image analysis formats.

Images can be uploaded either from local storage or Oracle Cloud Infrastructure Object Storage. The images can be in the following formats:
  • JPG
  • PNG