V1.0.7 Release of VOTT Visual Object Tagging Tool
VOTT provides end to end support for generating datasets for and validating object detection models.
- Computer assisted tagging of objects using a custom implementation of the Camshift algorithm
- Export Tags to CNTK-FastRCNN or YOLO format for training a object recognition model.
- Running and validating a trained object recognition model on new videos to generate stronger models.
Change log
- Fixed bug related to Yolo export of image directories
- Added support for evaluating python and brainscript for cntk fastrcnn models
- Fixed bug with image path display
- Enabled copy and paste on OSX
- Enabled go to last untagged frame
- Fixed export complete notification bug
- Added Ogg video support
- Added custom anaconda3 path support