3D computer vision, made easy.
Annotate entire scenes in one go, instead of labeling frames one by one. Deploy custom computer vision models in days, not weeks.
Get StartedAnnotate entire scenes in one go, instead of labeling frames one by one. Deploy custom computer vision models in days, not weeks.
Get StartedAnnotate dense 3D reconstructions. By labeling entire scans in one go, data annotation can be done in hours and not weeks. No need to outsource.
We offer ready-made detection models for both 2D and 3D detection, which you can easily train and tune on your own machine or in the cloud. Or use your own.
Starting from collecting data with your iPhone or RealSense to calibrating sensors, the Stray toolkit takes the pain away from having to implement everything from scratch.
The Stray toolkit offers easy ways for collecting RGB and depth training data using your iPhone or Intel RealSense camera.
With only a few scans, you can create large datasets for training detection models.
Stray Studio lets you reconstruct RGB and depth data into dense 3D models. Annotate objects of interest with bounding boxes, rectangles, and keypoints.
The annotations are projected back to the original images, saving enormous amounts of time compared to labeling images individually.
Train custom 2D and 3D detection models using the Stray CLI and labels from Stray 3D Studio.
Trained models can be easily integrated into existing applications via the Stray Vision SDK.
Run 3D reconstruction and detection models through a simple command line tool.
Integrate detection capabilities into your Python code.
Run reconstruction, train models, and store data in the cloud.
.svg){kind=small}
