### Install dependencies

Source: https://github.com/wisconsinaivision/universalfakedetect/blob/main/README.md

Install the required PyTorch and Torchvision libraries.

```bash
pip install torch torchvision
```

--------------------------------

### Clone the repository

Source: https://github.com/wisconsinaivision/universalfakedetect/blob/main/README.md

Initial steps to download the project source code and navigate into the directory.

```bash
git clone https://github.com/Yuheng-Li/UniversalFakeDetect
cd UniversalFakeDetect
```

--------------------------------

### Train Model

Source: https://github.com/wisconsinaivision/universalfakedetect/blob/main/README.md

Execute the training script with the required backbone configuration.

```bash
python train.py --name=clip_vitl14 --wang2020_data_path=datasets/ --data_mode=wang2020  --arch=CLIP:ViT-L/14  --fix_backbone
```

--------------------------------

### Dataset directory structure

Source: https://github.com/wisconsinaivision/universalfakedetect/blob/main/README.md

Expected directory layout for the test datasets.

```text
datasets
└── test					
      ├── progan	
      │── cyclegan   	
      │── biggan
      │      .
      │      .
```

--------------------------------

### Dataset Directory Structure

Source: https://github.com/wisconsinaivision/universalfakedetect/blob/main/README.md

Expected directory structure for the training dataset.

```text
datasets
└── train			
      └── progan			
           ├── airplane
           │── bird
           │── boat
           │      .
           │      .
```

--------------------------------

### Cite the research paper

Source: https://github.com/wisconsinaivision/universalfakedetect/blob/main/README.md

Use this BibTeX entry to cite the project in academic publications.

```bibtex
@inproceedings{ojha2023fakedetect,
      title={Towards Universal Fake Image Detectors that Generalize Across Generative Models}, 
      author={Ojha, Utkarsh and Li, Yuheng and Lee, Yong Jae},
      booktitle={CVPR},
      year={2023},
}
```

--------------------------------

### Evaluate Model on Datasets

Source: https://github.com/wisconsinaivision/universalfakedetect/blob/main/README.md

Run evaluation on all datasets or specific generative model paths using the validate.py script.

```bash
python validate.py  --arch=CLIP:ViT-L/14   --ckpt=pretrained_weights/fc_weights.pth   --result_folder=clip_vitl14
```

```bash
python validate.py  --arch=CLIP:ViT-L/14   --ckpt=pretrained_weights/fc_weights.pth   --result_folder=clip_vitl14  --real_path datasets/test/progan/0_real --fake_path datasets/test/progan/1_fake
```

=== COMPLETE CONTENT === This response contains all available snippets from this library. No additional content exists. Do not make further requests.