Release 2025/1/8
Last updated
Last updated
We’re kicking off the new year with exciting updates in DataVerse! The January release introduces enhanced model conversion capabilities and support for a new project type, further expanding the platform's versatility and functionality.
Model Convert with L4, L40 TensorRT Support:
The model conversion feature has been upgraded to include support for L4, L40 TensorRT conversion. This addition enhances compatibility and optimization, especially for environments requiring high-performance real-time detection.
VQA Project Support:
DataVerse now supports Visual Question Answering (VQA) project types. You can:
Create VQA Project:
Easily set up a new VQA project to organize your Visual Question Answering workflows. Customize project settings to suit your dataset and annotation requirements.
Import VQA Datasets:
Seamlessly upload VQA datasets into the platform, ensuring smooth integration of data for further processing and labeling.
Data Visualization:
Utilize advanced search tools and metrics to locate specific data points quickly. Manage datasets with intuitive filters and organization options, streamlining your VQA workflows.
VQA Labeling Panel:
An interactive annotation panel tailored for VQA tasks, allowing you to label questions and answers efficiently. Includes tools to manage and review annotations for better accuracy and collaboration.
VQA Metrics Calculation:
Future updates will include methods for calculating VQA metrics, providing detailed performance evaluation for VQA tasks. This feature will help users measure and optimize the effectiveness of VQA models.
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.
