Auto segmentation tracking (BETA)



The auto-segmentation tracking feature is currently available in Beta. While in Beta, the feature might not be as stable as our fully released functionalities, and we encourage users to provide feedback to help us improve its performance and reliability.

Auto-segmentation tracking follows instance labels through a series of frames in a video, allowing you to create near pixel perfect annotations with little effort. Auto-segmentation tracking is designed to maintain continuity in object tracking, even if the objects momentarily vanish from view for several frames.

Since tracking algorithms propagate labels forward through frames, auto-segmentation tracking works if labels are made on lower number frame (for example, near the start of the video).


The following table shows the Ontology shapes that the auto-segmentation tracking algorithm supports. Shapes not supported by the auto-segmentation tracking use the standard object tracking algorithm instead.



Auto-segmentation tracking is only available for videos.

Ontology shapeAuto-segmentation tracking
Bounding box
Rotatable bounding box

Using auto-segmentation tracking



Auto-segmentation tracking is computationally intensive and performance might be slower than other labelling actions.

  1. Create at least one bitmask or polygon instance label on the frame you want to start tracking the instance from.

  2. Click the icon on the instance to be tracked, and click Track from this frame to run auto-segmentation tracking for 30 consecutive frames. Initiate tracking on up to 20 instances.



Label Editor V2 allows you to right-click the shape, and run auto-segmentation tracking by clicking Run tracking for 1 object.

Tracking range

The range auto segmentation tracking runs can be adjusted in the Object tracking section of the editor settings.

The default tracking range is set to 30 frames. This range includes the frame that auto-segmentation tracking starts on.



Increasing the tracking range to more than 500 frames can slow down performance considerably. We recommend running auto-segmentation tracking in intervals of 500 frames at a time while this feature is in the beta phase. This feature's performance is continuously being improved, and these limits will be raised in the near future.