Trescope is a comprehensive 3D machine learning development tool devoted to improve developing experience and speed in 3D field, which helps researchers and developers to label, debug, visualize various 3D data . see the documentation for details .
Support labeling multiple attributes of 3D data, such as enumeration attributes, pose(coming soon), point cloud segmentation(coming soon) and so on
Enumeration attributes labeling example
Support breakpoint and 3D data visualization when you debug your program
Step-by-step debug example
Support visualization of various 3D data types, such as mesh, voxel, point cloud and so on, it is worth mentioning that we also support 3D-FRONT dataset visualization(which is a fantastic scene dataset from Alibaba)
3D-FRONT dataset visualization example
Support multiple development languages, and it is easy to support a new language under trescope architecture
Support visualize 3D data by two way:
- Online mode(display output): you can visualize 3D data in browser with interaction(zoom, rotate...)
- Offline mode(file output): you can visualize 3D data to file, in this way, you can check plenty of 3D data quickly
If you have any suggestion or issue, you can contact with us by Dingtalk or issue at github directly, scan the QR code below to enter Dingtalk group:
MIT
Trescope uses a number of open source projects to work properly: