thuyngch/image-forgery-using-deep-learning

Image Forgery Detection using Deep Learning, implemented in PyTorch.

Image-Forgery-using-Deep-Learning

Image Forgery Detection using Deep Learning, implemented in PyTorch.

Proposal

The whole framework: An RGB image, firstly, is divided into overlapping patches (64x64). Then, RGB patches are converted to the YCrCb color channel, before being scored by a network. Lastly, a post-processing stage is designed to refine predictions of the network and make a final conclusion on the authentication of the image.

The deep neural network is adapted from MobileNet-V2. However, we modify the original MobileNet-V2 to be more relevant to our problem. The picture below depicts the architecture modification.

Experimental results

We have conducted a comprehensive evaluation on model configurations to show which factor improves the final performance of the model. To figure out this, we define six configurations accompanied with the MobileNetV2, denoted as MBN2, as the core. There are two color channels to be considered, namely RGB and YCrCb. Besides, three MobileNetV2 architectures are taken into account for comparing. The first architecture is MobileNetV2 trained from scratch, the second one is MobileNetV2 initialized with pre-trained weights from ImageNet, and the last one is modified MobileNetV2 trained from scratch.

Citation

If you find this work useful, please cite:

@article{
  title={Preserving Spatial Information to Enhance Performance of Image Forgery Classification},
  author={Hanh Phan-Xuan, Thuong Le-Tien, Thuy Nguyen-Chinh, Thien Do-Tieu, Qui Nguyen-Van, and Tuan Nguyen-Thanh},
  journal={International Conference on Advanced Technologies for Communications (ATC)},
  year={2019}
}

Contact

• Nguyen Chinh Thuy
• Do Tieu Thien
• Nguyen Van Qui

image-forgery-using-deep-learning's Issues

scanned document Forgery

Hi,
Thanks for great work . i want to detect copy-move Forgery from scanned document like invoice or receipts . Is this code work for my problem if i trained model for invoice forgery ?

Thanks

Foregeryb detection test

Hello,

I want to test this project but I do not know how to do it (I'm beginanate), you can help me please? I consulted the technical report but it is empty.

thanks in advance

Getting different results in different systems

Hi,i have tried to run your code in my own pc and i am getting some score map results which is different when i try to run samething on GCP instance.I don't know whats the reason can you clarify that?

i have no idea about Positive and Negative

@AntiAegis , i have no idea about below code, and i don't know how to predict a new image tampered?
if decision==1:
if GT=="Positive":
print("Prediction: Positive ==> True")
else:
print("Prediction: Positive ==> False")
else:
if GT=="Positive":
print("Prediction: Negative ==> False")
else:
print("Prediction: Negative ==> True")

Execution problem

Hello,

Thank you for sharing your code that I find your work very interesting.

can you give more details on the execution of this code? how can I implement it? to test an image?

thanks in advance

about the model

Hi,
i tried the model with demo/compare-prediction.ipynb use our tamper image data . but the result is not good .
can you help me how to fineturn with your model in the project

Error in runing train_MBN2_mod.py

Hi.
I used CASIA 2.0 dataset and train MBN2 model, in running train_MBN2_mod.py, the system raised the error as below:

---

Train on epoch 1
0%| | 0/19 [00:00<?, ?it/s]
Traceback (most recent call last):
File "train_MBN2_mod.py", line 163, in
loss_train, acc_train, time_train = train_on_epoch(
File "/home/rr-ubuntu/pythonproject/Image-Forgery-using-Deep-Learning/utils/learning.py", line 32, in train_on_epoch
logits = model(X)
File "/home/rr-ubuntu/.local/lib/python3.8/site-packages/torch/nn/modules/module.py", line 727, in _call_impl
result = self.forward(*input, **kwargs)
File "/home/rr-ubuntu/pythonproject/Image-Forgery-using-Deep-Learning/utils/models.py", line 217, in forward
out = self.linear(out)
File "/home/rr-ubuntu/.local/lib/python3.8/site-packages/torch/nn/modules/module.py", line 727, in _call_impl
result = self.forward(*input, **kwargs)
File "/home/rr-ubuntu/.local/lib/python3.8/site-packages/torch/nn/modules/linear.py", line 93, in forward
return F.linear(input, self.weight, self.bias)
File "/home/rr-ubuntu/.local/lib/python3.8/site-packages/torch/nn/functional.py", line 1690, in linear
ret = torch.addmm(bias, input, weight.t())
RuntimeError: mat1 dim 1 must match mat2 dim 0

How can I do for solving this problem?

"Prediction: Positive ==> True"

Hi,
i run the demo/compare-prediction.ipynb and i have no idea about

like "Prediction: Positive ==> True" it means the pic is authentic?
and "Prediction: Positive ==> False" it means the pic is tampering?

call for help

can you give us the link to your paper about this project

No module named 'libs'

I got this error when run file prepare_data.py :

Traceback (most recent call last):
  File "/home/atsg/PycharmProjects/research/potential/Image-Forgery-using-Deep-Learning/utils/patches.py", line 16, in <module>
    from libs import image
ImportError: No module named 'libs'

I could not find any "libs" distribution in Pypi . Can you help me?
Anw, Could you give a brief tutorial how to run your project?
Thanks a lot.