Release 2024/09/18
Last updated
Last updated
We are excited to announce the release of several powerful new features and updates to the DataVerse Service, designed to enhance performance and functionality for all users.
Expanded YOLOv9 Model Types:
DataVerse now offers more flexibility with additional model types for YOLOv9. You can choose between c, e, and s variants, allowing you to tailor the model selection based on your specific requirements for accuracy and performance.
Frame/Sequence Data Slice:
A new feature now allows you to save data slices of frames or sequences. This provides more granular control over data management, making it easier to access and work with specific portions of your dataset for labeling, training and analysis.
Considering the characteristics of frame and sequence data slices, each frame or sequence will be packaged as a group of files and assigned to a labeler or reviewer. As such, the labeling panel for sequence task will have sensor selector and page switcher at the bottom of the panel.
you can find more information in the section of labeling/reviewing panel.
Upcoming Model Convert Feature:
Soon, you will be able to automatically convert models into TensorRT and other optimized versions. This will significantly enhance inference performance, ensuring faster and more efficient deployment in real-time environments.
Be sure to follow us for more details and updates as we approach the release date.
For technical support, please contact our support team during business hours.
