ColFigPhotoAttnNet

This project contains the ColFigPhotoAttnNet framework for reliable finger photo presentation attack detection, leveraging window-attention mechanisms across multiple color spaces. The architecture integrates MobileNet-V3 for feature extraction and applies pointwise convolution within a bottleneck framework with window attention mechanisms, using fine-tuned Swin transformer weights. Then, features of three color spaces are combined with element-wise addition and pointwise convolution and fed in a Nested Residual Block that has been initialized with ResNet34 weights. Finally, at inference, the model applies Dynamic Quantization and gives the final global decision.

Github Link

@inproceedings{colfigphotoattnnet,
  title={ColFigPhotoAttnNet: Reliable Finger Photo Presentation Attack Detection Leveraging Window-Attention on Color Spaces},
  author={Anudeep Vurity, Emanuela Marasco, Raghavendra Ramachandra, Jongwoo Park},
  booktitle={IEEE Winter Applications of Computer Vision (WACV)},
  year={2025},
  pages={1--10}
}