About Project
User friendly semi-automation based Annotation Tool to label video and image data with multiple annotation types using Computer vision algorithms, for generating data to train ML/DL models.
Client
HCL Technologies Limited
Category
Frontend Development
Key Highlights
- Optimized the tool to improve annotation quality for creating datasets to train Deep Learning models.
- Worked on Multiple Object tracking algorithm for annotating video data with semi-automation.
- Added features to support productivity measurement of users (annotators).
LIBRARIES AND TOOLS
- PyQT for UI
- Pandas for productivity measurement
- OpenCV for image/video processing and labelling
- Dlib for multi-object tracking
- Traffic video data captured