Fast and Easy Data Annotation

 

Get maximum speed without sacrificing quality. A desktop application with a UX built for through-put, HyperLabel is a complete toolset for quality labeling process management and training data creation.

Free For An Unlimited Quantity Of Labels

HyperLabel Developer provides:

  • Complete ownership, privacy and control of video and image data
  • Fast 5-step process to high-quality ML labeling
  • Easy-to-define custom schemas 
  • Flexible import/export speeds time-to-training

 

5 Simple Steps To Labeling High-Quality Training Datasets

Name your project and select data sources. Your data sources can include both images and video. Video is added as frame-by-frame images. Choose how many frames-per-second to capture.

 

Connect data sources from cloud storage, local storage, or local CSV to import datasets into your project.

Create your own custom schema by defining the labels you will use to label data with different label types to support your ML models.

 

Data label types include: Rectangle, Polygon, Point, Feature Points, Text, Select, and Multi Select.

Easy-to-use Labeler UI lets you label objects quickly without unnecessary clicks. Quick image handling tools and copy/paste of labels from prior images helps speed things up even more.

 

Always know how much you have left in the queue and if you’ve already labeled the file you are viewing.

Review project labels before you export them for use in your ML projects. 

 

Monitor labeling status, labeler productivity, and use filtering options to view the project information the way you want–by labeler, data source, status, or data id.

Output your labels and files in a standard way to import and use them in your ML projects. 

 

Export to JSON, or export image classifications; object detections in Create ML, YOLO or COCO; feature points in COCO, semantic segmentations in COCO. Select if you want to include image files as well.

Easy-To-Define Custom Schemas

Rectangles

Draw boxes to detect specific objects within images.

Polygons

Segment objects for detection within image frames.

Point

Label anatomical or structural points of interest within images.

Feature Points

Use for highly detailed detection such as emotion & body pose.

Text

Leverage for free text production for image captions ideal for accessibility.

Select

Designed for specific filters & image search.

Multi Select

Apply for multi label image classification & attribute identification.

Download today to start on the fastest path to machine learning.

 

HyperLabel Enterprise Tailors Custom Installations To Meet The Needs Of Your Business

HyperLabel Enterprise provides:

  • Flexible integration based on each customer’s needs
  • ML workflow process integration
  • On-premise installation and dedicated support

Coming Soon: HyperLabel Pro

HyperLabel Pro Is Ideal For Data Science and IT Teams

HyperLabel Pro provides:

  • Cloud projects collaboration
  • Deep learning powered object tracking
  • 3D data processing

Unrestricted Labeling

Scale and sync projects with quality-enforced queueing and data access while you manage team expertise, roles and performance.

Cloud Projects Collaboration

Scale and sync projects with quality-enforced queueing and data access while you manage team expertise, roles and performance.

ML Assisted Labeling

The HyperLabel model and pre-trained classifiers enable predictive labeling of common objects. Reduce time and increase accuracy.

QA Interface

HyperLabel’s QA Interface enables quick and efficient batch reviews of labeled data before using it for training.

Deep Learning Object Tracking

Hyper-accelerate labeling by 50x. First-frame labeling and label persistence for subsequent adjustments.

3D Data Processing

Flexibly import 3D file formats including LiDAR, DICOM (MRI or Ultrasound), as well as mapping and satellite imagery.