CoReD: Generalizing Fake Media Detection with Continual Representation using Distillation (ACMMM'21 Oral Paper)
(Accepted for oral presentation at ACMMM '21)
Paper Link: (arXiv) (ACMMM version)
## Overview
We propose _Continual Representation using Distillation_ (_CoReD_) method that employs the concept of **Continual Learning (CL)**, **Representation Learning (RL)**, and **Knowledge Distillation (KD)**.
### Comparison Baselines
- Transfer-Learning (_TL_) : The first method is Transfer learning, where we perform fine-tuning on the model to learning the new Task.
- Distillaion Loss (_DL_) : The third method is a part of our ablation study, wherewe only use the distillation loss component from our CoReD loss function to perform incremental learning.
- Transferable GAN-generated Images Detection Framewor (_TG_) : The second method is a KD-based GAN image detection framework using L2-SP and self-training.
## Training & Evaluation
### - Requirements and Installation
We recommend the installation using the _requilrements.txt_ contained in this Github.
python==3.8.0
torchvision==0.9.1
torch==1.8.1
sklearn
numpy
opencv-python
```console
pip install -r requirements.txt
```
####
- **Note that** :
## - Train & Evaluation
### - Full Usages
```console
-m Model name = [CoReD, KD, TG, FT]
-n Network name = [Xception, Efficient']
-te Turn on test mode [True, False]
--name_sources Name of 'Source' datasets. one or multiple names. (ex. DeepFake / DeepFake_Face2Face / DeepFake_Face2Face_FaceSwap)
--name_target Name of 'Target' dataset. only a single name. (ex.DeepFake / Face2Face / FaceSwap / NeuralTextures) / used for Train only')
--data Dataset path. it must be contained Sources & Target folder name
--weigiht You can select the full path or folder path included in the '.pth' file
-lr Learning late (For training)
-a Alpha of KD-Loss
-nc Number of Classes
-ns Number of Stores
-me Number of Epoch (For training)
-nb Batch-Size
-ng GPU-device can be set as ei 0,1,2 for multi-GPU (default=0)
```
####
### - Train
To train and evaluate the model(s) in the paper, run this command:
- **Task1**
We must train pre-trained single model for task1 .
```TRAIN
python main.py -t={Source Name} -d={folder_path} -w={weights}
python main.py -t=DeepFake -d=./mydrive/dataset/' #Example
```
- **Task2 - 4**
```TRAIN
python main.py -s={Source Name} -t={Target Name} -d={folder_path} -w={weights}
python main.py -s=Face2Face_DeepFake -t=FaceSwap -d=./mydrive/dataset/ -w=./weights' #Example
```
- **Note that** If you set _-s=Face2Face_DeepFake -t=FaceSwap -d=./mydrive/dataset -w=./weights_ when you start training, data path **"./mydrive/dataset"** must include **'Face2Face', 'DeepFake', and 'FaceSwap'**, and these must be contained the **'train','val'** folder which include **'real'&'fake'** folders.
### - Evaluation
After train the model, you can evaluate the dataset.
- **Eval**
```EVAL
python main.py -d={dataset as full name} -w={weights} --test
python main.py -d=./mydrive/dataset/DeepFake/testset -w=./weights/bestmodel.pth --test #Example
```
## - Result
- **AUC scores (%)** of various methods on compared datasets.
#### - Task1 (GAN datasets and FaceForensics++ datasets)
