What it does
This is the third part of a series on vector databases for AI agents. It takes an image as input and compares it with a reference dataset already uploaded to Qdrant, using stored class centres and thresholds to decide whether the image is an anomaly. It works with any image dataset, but the earlier setup flows must be run first.
Use cases
- 01Flag images that differ from a reference set
- 02Screen incoming photos before they are filed
- 03Test anomaly detection on a custom image dataset
Connects
HTTP Request · Code