Features
VisionCloud is a comprehensive platform designed to streamline the process of creating, training, testing, and deploying deep learning models for industrial machine vision applications.
Wondering what makes VisionCloud stand out? Here are some of its key features:
Image management
✓ Easily upload images from your project and organize them into datasets.
✓ You can create multiple datasets within a project to categorize your images based on different criteria, such as image type, source, or labeling status.
✓ If your project gets too big, datasets can also be organized in folders.
✓ All datasets are stored in a centralized location, making them easy to manage and access by your team.
Advanced labeling tools
✓ VisionCloud provides powerful labeling tools to quickly annotate your images. You can create bounding boxes, polygons, and add classes to images.
✓ Each image has a labeling history, thus enabling you to track changes, know who made them, and revert to previous versions if needed.
Version control & history
✓ Keep track of changes to your datasets and models with built-in version control. Access previous versions, compare changes, and revert if necessary.
✓ All version management is done by the platform automatically, so you can focus on your work while the changes are tracked in the background.
Team collaboration
✓ Working with colleagues on the same project often requires sharing large datasets, which can be cumbersome. VisionCloud makes team collaboration easy.
✓ Invite team members to your projects, assign roles, and collaborate seamlessly on data labeling and model training.
✓ Multiple users can label the same dataset concurrently, streamlining the annotation process. You can add comments and discuss specific images or annotations directly within the platform.
Train directly in the cloud
✓ VisionCloud allows you to train deep learning models directly on the platform using your labeled data, eliminating the need for local infrastructure and simplifying the training process. ✓ You can also launch multiple training sessions in parallel, making it easy to experiment with different model architectures and parameters.
Model exporting
✓ Once your model is trained, you can export it in ONNX format for deployment in various environments, including integration with OneVision.
✓ With OneVision, you can deploy your trained models for inference in production environments, connect with cameras, communicate with PLCs, use classic machine vision tools, and more.
🔗 Learn more about OneVision in the OneVision documentation.