voqtuyen / deep-pix-bis-pad.pytorch

An unofficial implementation of the paper "Deep Pixel-Wise Binary Supervision for Face Presentation Attack Detection" in Pytorch

Hi there 👋

I am Tuyen,

• 🔭 I’m currently working on machine learning, backend, data analysis
• 🌱 I’m currently learning data engineering
• 👯 I’m looking to collaborate on machine learning and data related fields
• 💬 Ask me about anything :)
• 📫 How to reach me: [email protected]
• 🌀 Mindset: open, flexible and relative
• ⚡ Fun fact: :D

deep-pix-bis-pad.pytorch's Issues

Request for dataset

Can you provide the link to CASIA-SURF dataset or NUAA dataset?
Both the original links are broken

Did you get problem running with GPU?

I tried to run your code but it stucked at line 13 of Trainer.py self.network = self.network.to(device), it said RuntimeError: cuda runtime error (100) : no CUDA-capable device is detected at ..\aten\src\THC\THCGeneral.cpp:50

Any special config? I'm currently using pytorch 1.4.0

ModuleNotFoundError: No module named 'models.detector'

Running on python3.10 virtual environment on windows 11.

Please how do i resolve this Issue, where do i get the missing module?

DeePixBiS convertation to rknn

I am trying to convert the DeePixBiS model, but on the inference I get a completely different result in contrast to the original implementation of the model. This is how I convert from onnx to rknn https://pastebin.com/cNJvijCe , and this is the inference code https://pastebin.com/jy6Y6EeE . Could you tell me what I 'm doing wrong ?