Find objects in an image at a pixel level.
Segment uses a zero-shot model which means you can specify any class you like.
The model behind Segment will be downloaded automatically the first time you use it on your computer. This can take a few minutes depending on your internet connection.
class- The class to detect. If no class is specified, all objects will be segmented.
The following example loads an image, runs inference to find all objects matching the "person" class, and displays the results.
Load["./cats.jpeg"] Segment["cat"] Show