1 software-driven augmentation tool for converting standard facial images into masked versions. It applies synthetic overlays to existing face datasets to create training data for masked face recognition systems.
Use Cases
- Train a mask-detection classifier using the generated masked images
- Improve facial recognition performance by training on images with synthetic mask overlays
- Evaluate biometric system error rates using the masked face output images
Strengths
- Supports synthetic mask overlays on existing facial image files
- Uses facial landmark detection to ensure accurate mask placement over the nose and mouth
- Provides a mechanism to convert standard face datasets into masked versions for benchmarking