Migrate from Ground Truth Labeler and Lidar Labeler Apps to Multi-Sensor Labeler App

R2026b

Why Migrate to Multi-Sensor Labeler App?

Starting in R2026b, the Multi-Sensor Labeler app replaces the Ground Truth Labeler and Lidar Labeler apps with a single, unified labeling environment. With the Multi-Sensor Labeler, you can:

  • Label images, videos, and point clouds simultaneously in one app

  • Store ground truth labels using the groundTruthMultiSensor object and create label definitions using the labelDefinitionCreatorMultiSensor object

  • Use consistent labeling tools and workflows across multiple sensor types

  • Use specialized tools for point cloud annotation such as semantic point cloud labeling

  • Use semi-automated algorithms to make 3D labeling easier such as Burst mode

  • Use fully automated tracking and object detection algorithms for point clouds

  • Export labeled annotation data and create training data directly to train deep learning models

Note

The Ground Truth Labeler and Lidar Labeler apps will be removed in a future release and are no longer recommended for new labeling projects.

Key Changes in Multi-Sensor Labeler app

The following table summarizes the key changes from the Ground Truth Labeler and Lidar Labeler apps to the Multi-Sensor Labeler app.

CategoryGround Truth Labeler / Lidar LabelerMulti-Sensor Labeler
App commandgroundTruthLabeler / lidarLabelermultiSensorLabeler
Ground truth objectgroundTruthMultisignal (Automated Driving Toolbox) / groundTruthLidargroundTruthMultiSensor
Label definition creatorlabelDefinitionCreatorMultisignal / labelDefinitionCreatorLidarlabelDefinitionCreatorMultiSensor
Connector classdriving.connector.Connector / lidar.syncImageViewer.SyncImageViewerlidar.connector.Connector

Open Multi-Sensor Labeler App

You can open the Multi-Sensor Labeler app in two ways:

  • Command line — Type multiSensorLabeler at the MATLAB command prompt.

  • Apps tab — On the MATLAB toolstrip, click the Apps tab, then click Multi-Sensor Labeler in the app gallery.

The multiSensorLabeler command supports the same calling syntaxes as the previous groundTruthLabeler and lidarLabeler commands:

% Open the app with no data
multiSensorLabeler

% Open with a video
multiSensorLabeler(videoFile)

% Open with an image sequence and timestamps
multiSensorLabeler(imageFolder, timestamps)

% Open with a point cloud sequence
multiSensorLabeler(ptCloudSeq)

% Open with a session file
multiSensorLabeler(sessionFile)

% Open with a ground truth object
multiSensorLabeler(gTruth)

% Open with a connector
multiSensorLabeler('ConnectorTargetHandle', connectorObj)

Use Your Existing Scripts and Ground Truth Objects

Your existing scripts and ground truth objects continue to work with the Multi-Sensor Labeler app.

  • Ground truth objects — You can load groundTruthMultisignal (Automated Driving Toolbox) objects from the Ground Truth Labeler app and the groundTruthLidar objects from the Lidar Labeler app directly into the Multi-Sensor Labeler. The app accepts these objects and loads their label definitions and label data automatically.

    Note

    Support for loading legacy groundTruthMultisignal and groundTruthLidar objects will be removed in a future release. To preserve your work, load these objects in the Multi-Sensor Labeler app and export them as groundTruthMultiSensor objects.

  • Automation algorithms — If you have custom automation algorithms in +vision/+labeler/ or +lidar/+labeler/ package folders, move them to +pointcloud/+labeler/ so the Multi-Sensor Labeler can discover and use them.

  • Label definition files — MAT files containing label definitions exported from the Ground Truth Labeler or Lidar Labeler can be imported directly into the Multi-Sensor Labeler.

Open Sessions from Ground Truth Labeler or Lidar Labeler Apps in Multi-Sensor Labeler App

You can open existing session files from both legacy apps in the Multi-Sensor Labeler app:

  • Ground Truth Labeler sessions — Load .prj or .mat project files. All label definitions, labeled data, and session settings are preserved.

  • Lidar Labeler sessions — Load .mat session files. All label definitions (including Voxel/Semantic Point ROI labels), labeled data, and session settings are preserved.

Note

Support for loading legacy Ground Truth Labeler and Lidar Labeler session files will be removed in a future release. To preserve your work, open the legacy sessions in the Multi-Sensor Labeler app and save them as a new project file.

To load a legacy session: 

% Load a Ground Truth Labeler project file
multiSensorLabeler('myGTLproject.prj')

% Load a Lidar Labeler session
multiSensorLabeler('myLLsession.mat')

See Also

| |

Topics