# Vision service configuration

The vision service has three built-in models with distinct configuration shapes:

**mlmodel**  
Configure the mlmodel vision service to turn a deployed ML model into a detector, classifier, or 3D segmenter.

**color_detector**  
Configure the color_detector vision service to find regions of a specific hue in camera images. No ML model required.

**detections-to-segments**  
Project 2D detections into 3D point cloud segments using a depth camera's intrinsic parameters.

For API method reference (`GetDetections`, `GetClassifications`, `GetObjectPointClouds`, `CaptureAllFromCamera`, `GetProperties`), see [Vision service API](https://docs.viam.com/reference/apis/services/vision/).