A wealth of powerful features awaiting your use
Our MaxusAI LabelEngine has all the features you’d expect to make your data labelling efforts a breeze.
AI assistance
as You Go
Active Learning (Pro)
Let the AI predict labels from just less than a hundred samples. Accept or refine the predicted labels. The AI learns from it and will improve much faster.
Auto-Segmentation (Pro)
Let the AI predict the polygon masks for you. Create segmentation masks faster and easier by just defining 4 points - top, bottom, most left and most right.
Active Detection (Pro)
Let the AI scan your images and videos to detect potential candidates in minority classes, allowing you to scale from a few small samples
Create high
Quality Datasets
All Polygons
As opposed to bounding boxes, polygons or pixel masks are your objects already without noisy background.
Images & Videos
Annotate Images & Video frames with ease. Use the powerful preview pane to get a glance of your progress, verified assets and yet to do.
Advanced Labeleditor
Setup domain or industry specific labels with ease and extend as you go. Manage hierarchies of main objects with their attributes and conditions.
Scale With
Your needs
Label Analytics
Monitor the distribution of your annotations to manage and avoid class imbalances. Let AI-Active Labelling (Pro) detect potential minority observations.
Distribute Work (Pro)
Sharing is caring. Create projects and invite team members. Share the workload, and track progress and finish faster.
No Lock-ins
Your work belongs to you. We will support the export of your annotations into COCO, YOLO, PASCAL VOC and the download of your models.
Get Started
Within 2 Minutes
In-Browser
Works in any modern browser. No advanced installation or configuration required, e.g. via node or docker.
Local & High Quality
Annotate original assets from your local folders and disks. No lengthy uploads required. No additional compression of your images or videos.
Secure
We only store metadata of your assets to match progress within projects. Though, AI-Assistance requires that lower resolution data is sent to us.