pei-kaihuang / icip23_d-ldcformer

ICIP'23 Incorporating Learnable Descriptive Convolution to Vision Transformer for Face Anti-Spoofing

LDCformer

LDCformer: Incorporating Learnable Descriptive Convolution to Vision Transformer for Face Anti-Spoofing (ICIP '23)

Decoupled Learnable Descriptive Convolution (Decoupled-LDC)

Architecture of LDCformer

Architecture of Decoupled-LDC Block

Requirements

numpy==1.23.3
pytz==2022.4
requests==2.28.1
scikit_learn==1.2.0
timm==0.6.7
torch==1.10.1
torchvision==0.11.2

Training & Testing

Run train.py to train LDCformer

Run test.py to test LDCformer

Citation

If you use the LDCformer/Decoupled-LDC, please cite the paper:

@inproceedings{huang2023ldcformer,
 title={LDCformer: Incorporating Learnable Descriptive Convolution to Vision Transformer for Face Anti-Spoofing},
 author={Huang, Pei-Kai and Chiang, Cheng-Hsuan and Chong, Jun-Xiong and Chen, Tzu-Hsien and Ni, Hui-Yu and Hsu, Chiou-Ting},
 booktitle={2023 IEEE International Conference on Image Processing (ICIP)},
 pages={121--125},
 year={2023},
 organization={IEEE}
}
@inproceedings{huang2022learnable,
 title={Learnable Descriptive Convolutional Network for Face Anti-Spoofing},
 author={Huang, Pei-Kai and H.Y. Ni and Y.Q. Ni and C.T. Hsu},
 booktitle={BMVC},
 year={2022}
}

icip23_d-ldcformer's Issues

dataset format is .npy, cannot run

hello @Pei-KaiHuang , this is a nice work.
I see you load data using .npy format files(in train.py line 77). Can you give me more detail about the .npy format files?
Or can you provide a method to directly load images as a dataset?