13,000 annotated images from the University of Oxford Hand Dataset curated for hand detection and tracking. The collection includes bounding box annotations for hands across diverse video sequences and environments to support Faster RCNN model training within the Tensorflow Object Detection API.
Use Cases
- Train a hand localization model using bounding box coordinates to detect hand positions in static images
- Implement a temporal hand tracking algorithm by analyzing hand movement across sequential video frames
- Evaluate the inference speed and accuracy of Faster RCNN models on the specific domain of human hand detection
Strengths
- Includes bounding box annotations for hand localization across various video frames
- Sourced from the University of Oxford's specialized Hand Dataset
- Configured for immediate integration with the Tensorflow Object Detection API
- Supports Faster RCNN model implementations for high-accuracy detection