Video annotation for computer vision

Lodestar provides the world’s first active learning video annotation platform to accelerate high-quality dataset and model creation.

Video annotation suite for computer vision

Trained AI model on day one. 

Faster labeling

Annotate 5x times faster using custom predictive labeling and AI-powered frame selection. 

Faster time to a model

Continuous model training and a shared, managed dataset allow annotators and data scientists to collaborate and create a functional object detection model in an hour.

More video

Automated data preparation allows you to drag and drop hours of video into a project. Active learning automatically curates sparse datasets for you.

Active learning for video annotation

Active learning improves the dataset faster.

Active learning ranks the most important frames for annotation. A custom AI model searches the entire video to find the most valuable objects that will increase its accuracy.

More labels faster with custom predictive labeling.

Stop working with a generic prediction model. Label just 20 objects and a custom model, trained on your data, automatically starts labeling for you. Make a few adjustments to the machine generated labels and move on. It makes labeling at least five times faster than manual labeling.

Custom predictive labeling
Dataset management

Annotate. Collaborate. Iterate.

Experiment directly on the dataset while your labelers are working in the annotation tool. Filter annotations down to the pixel level, train different inference models, and create reports with integrated Jupyter notebooks. Stop wasting time creating and keeping track of dataset copies and manually slicing and dicing your data every time you run an experiment – work on a single source of truth.

Experts drive high-quality results.

Model training is constantly under control. Human experts provide real-time feedback to the machine on its predictive labels. The machine responds with continuous training cycles to improve the custom model. This human-in-the-loop system improves the predictive labeling and active learning.

Human-in-the-loop

On average 4x shorter time to model

Today, training a computer vision model involves extensive data preparation and curation before labeling begins. Data scientists wait weeks for the dataset to be labeled before they can train and validate their model. Typical model development takes multiple cycles before achieving model goals. 

Traditional computer vision model training process
Computer vision model training process using active learning

Lodestar changes all that. An integrated platform for labeling and training with native video support trains a model in real time and allows your teams to work in parallel. Active learning and AI prediction help generate labels and automatically curate your data. You’ll finish in weeks what used to take months.

One stop shop

Annotation suite

Everything you need to label a high-quality video dataset. This is where you’ll set up new projects, upload video, and monitor progress. Once a project is ready, you can get subject matter experts started on the labeling simply by sharing a link.

  • Easily navigate through hours of video with familiar player controls.
  • Zoom in and out of the frame to verify objects and refine bounding boxes.
  • Create object categories and monitor annotation counts without leaving the annotator.
  • Add notes to annotations.
Video and image annotation for computer vision
Deep learning platform for computer vision

Deep learning platform

Lodestar makes it easy to start a project and start labeling video in minutes. But our project management and integrated data science and IT resources make it a complete platform for developing computer vision models from video data.

DS Workbench gives you online access to the data and model your labelers are developing in Video Annotator using Jupyter.

IT Workbench includes the tools you need to seamlessly scale out and manage compute and storage resources.

Lodestar is a game changer.

“Typically we’d spend up to two months with a client on a proof of concept; labeling video data sets and developing a computer vision model to the point they were confident it would meet their needs and the project would be successful. Using Lodestar Navigator we’re getting there in a matter of hours. It’s not only accelerating our projects and increasing the return on investment, it’s changing the way we approach clients on these projects—we can spend an hour with them and demonstrate results instead of talking about possibilities.”

— Product Manager, IoT consulting firm

Book a demo

See how you can get results in under a day.