### Setup FFmpeg with H264 Codec

Source: https://github.com/ondyari/faceforensics/blob/master/dataset/README.md

Run this bash script to install FFmpeg along with the H264 codec, which is necessary for compression tasks.

```bash
bash setup_ffmpeg_h264.sh
```

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

### Install Faceswap Dependencies

Source: https://github.com/ondyari/faceforensics/blob/master/dataset/DeepFakes/faceswap-master/INSTALL.md

Run this command to install the project's Python dependencies. If it fails, manually install packages from requirements.txt.

```bash
python setup.py
```

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

### Successful Setup Log without Docker

Source: https://github.com/ondyari/faceforensics/blob/master/dataset/DeepFakes/faceswap-master/INSTALL.md

A log indicating a successful installation and dependency check for Faceswap when Docker is not enabled, but CUDA is. It highlights potential issues with TensorFlow compatibility for specific CUDA versions.

```text
INFO    The tool provides tips for installation
        and installs required python packages
INFO    Setup in Linux 4.14.39-1-MANJARO
INFO    Installed Python: 3.6.5 64bit
INFO    Installed PIP: 10.0.1
Enable  Docker? [Y/n] n
INFO    Docker Disabled
Enable  CUDA? [Y/n] 
INFO    CUDA Enabled
INFO    CUDA version: 9.1
INFO    cuDNN version: 7
WARNING Tensorflow has no official prebuild for CUDA 9.1 currently.
        To continue, You have to build your own tensorflow-gpu.
        Help: https://www.tensorflow.org/install/install_sources
Are System Dependencies met? [y/N] y
INFO    Installing Missing Python Packages...
INFO    Installing tensorflow-gpu
INFO    Installing pathlib==1.0.1
......
INFO    Installing tqdm
INFO    Installing matplotlib
INFO    All python3 dependencies are met.
        You are good to go.

```

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

### Get Faceswap Train Help

Source: https://github.com/ondyari/faceforensics/blob/master/dataset/DeepFakes/faceswap-master/USAGE.md

View all available arguments for the train command by using the help flag.

```bash
python faceswap.py train -h
```

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

### Get FaceSwap Help

Source: https://github.com/ondyari/faceforensics/blob/master/dataset/FaceSwapKowalski/README.md

Displays help information for the FaceSwap script, including available arguments and options.

```shell
python faecswap.py -h
```

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

### Install FaceSwap Requirements

Source: https://github.com/ondyari/faceforensics/blob/master/dataset/FaceSwapKowalski/README.md

Installs Python 2.7 dependencies from the requirements file.

```shell
pip install -r requirements.txt
```

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

### Get Faceswap Extract Help

Source: https://github.com/ondyari/faceforensics/blob/master/dataset/DeepFakes/faceswap-master/USAGE.md

View all available arguments for the extract command by using the help flag.

```bash
python faceswap.py extract -h
```

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

### Install Dependencies for DeepFakes

Source: https://github.com/ondyari/faceforensics/blob/master/dataset/DeepFakes/README.md

Installs Python 3.6, Cuda, Cudnn, nccl, and other necessary libraries for DeepFakes manipulation. Ensure you are on Ubuntu 16.04 with a compatible Nvidia GPU.

```shell
bash setup.sh
```

```shell
pip install -r faceswap-master/requirements.txt
```

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

### Build and Run Faceswap with Docker and CUDA

Source: https://github.com/ondyari/faceforensics/blob/master/dataset/DeepFakes/faceswap-master/INSTALL.md

Commands to build a Docker image for Faceswap with GPU support and run it. Includes options for running with or without a GUI, mounting volumes, and setting up display and audio/video permissions for the GUI.

```bash
docker build -t deepfakes-gpu -f Dockerfile.gpu .

```

```bash
# without gui. tools.py gui not working.
docker run -p 8888:8888 \
    --hostname faceswap-gpu --name faceswap-gpu \
    -v /opt/faceswap:/srv \
    faceswap-gpu

```

```bash
# with gui. tools.py gui working.
## enable local access to X11 server
xhost +local:
## enable nvidia device if working under bumblebee
echo ON > /proc/acpi/bbswitch
## create container 
nvidia-docker run -p 8888:8888 \
    --hostname faceswap-gpu --name faceswap-gpu \
    -v /opt/faceswap:/srv \
    -v /tmp/.X11-unix:/tmp/.X11-unix \
    -e DISPLAY=unix$DISPLAY \
    -e AUDIO_GID=`getent group audio | cut -d: -f3` \
    -e VIDEO_GID=`getent group video | cut -d: -f3` \
    -e GID=`id -g` \
    -e UID=`id -u` \
    faceswap-gpu

```

```bash
docker exec faceswap-gpu python /srv/tools.py gui

```

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

### Faceswap Convert Help

Source: https://github.com/ondyari/faceforensics/blob/master/dataset/DeepFakes/faceswap-master/USAGE.md

View the full list of arguments available for the convert command by using the help flag.

```bash
python faceswap.py convert -h
```

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

### Run Faceswap Tools

Source: https://github.com/ondyari/faceforensics/blob/master/dataset/DeepFakes/faceswap-master/INSTALL.md

Basic commands to run the Faceswap tools from the command line. Use the -h option for a list of available options.

```bash
python faceswap.py -h

```

```bash
python faceswap.py gui

```

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

### Download All FaceForensics++ Videos

Source: https://github.com/ondyari/faceforensics/blob/master/dataset/README.md

Use this command to download all light compressed original and manipulated videos for all three manipulation methods. Append `--num_videos <number>` to download a specific number of samples.

```python
python download-Faceforensics.py <output path> -d all -c c23 -t videos
```

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

### Run Faceswap GUI

Source: https://github.com/ondyari/faceforensics/blob/master/dataset/DeepFakes/faceswap-master/README.md

Launch the graphical user interface for Faceswap by running this command.

```bash
python faceswap.py gui
```

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

### Run Detection from Video

Source: https://github.com/ondyari/faceforensics/blob/master/classification/README.md

Execute the detection script on a single video file or a folder of videos. Specify input, model, and output paths. Enable CUDA with the --cuda flag or view all parameters using -h.

```shell
python detect_from_video.py
-i <path to input video or folder of videos with extenstion '.mp4' or '.avi'>
-m <path to model file, default is imagenet model
-o <path to output folder, will contain output video(s)
```

```shell
python detect_from_video.py -h
```

```shell
python detect_from_video.py --cuda
```

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

### Create DeepFakes Models

Source: https://github.com/ondyari/faceforensics/blob/master/dataset/DeepFakes/README.md

Generates DeepFakes models from image folders. Use `--keep_temp_directories` to retain intermediate files and `--convert_images` to create manipulated output images. Training requires a powerful GPU and can take approximately one day per model.

```shell
python deepfakes.py
    -i <path to image folders>
    -o <output path>
    --python_path <path to python file, e.g., /home/<user>/anaconda3/bin/python>
    -it 200000
    (--keep_temp_directories)
    (--convert_images)
```

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

### Train Model with Faceswap

Source: https://github.com/ondyari/faceforensics/blob/master/dataset/DeepFakes/faceswap-master/README.md

Execute this command to train a model using photos from two folders containing faces. The trained model is saved in the 'models' folder.

```bash
python faceswap.py train
```

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

### Download FaceForensics Script Usage

Source: https://github.com/ondyari/faceforensics/blob/master/dataset/README.md

This command demonstrates the general usage of the download script for the FaceForensics dataset. Use `python download-FaceForensics.py -h` for a full list of parameters.

```shell
python download-FaceForensics.py
    <output path>
    -d <dataset type, e.g., Face2Face, original or all>
    -c <compression quality, e.g., c23 or raw>
    -t <file type, e.g., videos, masks or models>
```

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

### Train Faceswap Model

Source: https://github.com/ondyari/faceforensics/blob/master/dataset/DeepFakes/faceswap-master/USAGE.md

Initiate the training process for the face swapping model. Specify the directories for the two faces and the model output.

```bash
python faceswap.py train -A ~/faceswap/data/trump -B ~/faceswap/data/cage -m ~/faceswap/models/
# or -p to show a preview
python faceswap.py train -A ~/faceswap/data/trump -B ~/faceswap/data/cage -m ~/faceswap/models/ -p
```

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

### Run FaceSwap Script

Source: https://github.com/ondyari/faceforensics/blob/master/dataset/FaceSwapKowalski/README.md

Executes the FaceSwap script to manipulate video frames. Requires source and target image folders.

```shell
python faceswap.py
    -i1 <source folder, images in format '04d.png'>
    -i2 <target folder, images in format '04d.png'>.
```

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

### Access Video Conversion Tool

Source: https://github.com/ondyari/faceforensics/blob/master/dataset/DeepFakes/faceswap-master/README.md

This command accesses the effmpeg conversion tool for videos. Alternatively, ffmpeg can be used for video-to-image and image-to-video conversions.

```bash
python tools.py effmpeg -h
```

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

### Convert Faces with Faceswap

Source: https://github.com/ondyari/faceforensics/blob/master/dataset/DeepFakes/faceswap-master/README.md

Use this command to apply new faces to photos in the 'original' folder, saving the modified images to the 'modified' folder.

```bash
python faceswap.py convert
```

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

### Download DeepFakeDetection Dataset Videos

Source: https://github.com/ondyari/faceforensics/blob/master/dataset/README.md

Obtain videos from the DeepFakeDetection dataset, specifying either the full dataset or original videos with raw compression.

```python
python download-FaceForensics.py <output path> -d <DeepFakeDetection or DeepFakeDetection_original>-c raw -t videos
```

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

### Extract Faces with Faceswap

Source: https://github.com/ondyari/faceforensics/blob/master/dataset/DeepFakes/faceswap-master/README.md

Run this command to extract faces from photos in the 'src' folder to the 'extract' folder.

```bash
python faceswap.py extract
```

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

### Convert Faces in Images

Source: https://github.com/ondyari/faceforensics/blob/master/dataset/DeepFakes/faceswap-master/USAGE.md

Use this command to convert faces in a directory of images using a trained model. Specify input, output, and model directories.

```bash
python faceswap.py convert -i ~/faceswap/photo/trump/ -o ~/faceswap/output/ -m ~/faceswap/models/
```

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

### Download Original Raw Compressed Videos

Source: https://github.com/ondyari/faceforensics/blob/master/dataset/README.md

Download all raw/lossless compressed original videos. This ensures no visual information is lost during compression.

```python
python download-FaceForensics.py <output path> -d original -c raw -t videos
```

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

### Extract Faces with Faceswap

Source: https://github.com/ondyari/faceforensics/blob/master/dataset/DeepFakes/faceswap-master/USAGE.md

Use this command to extract faces from input images and save them as training data. Specify input and output directories.

```bash
python faceswap.py extract -i ~/faceswap/photo/trump -o ~/faceswap/data/trump
# To convert cage:
python faceswap.py extract -i ~/faceswap/photo/cage -o ~/faceswap/data/cage
```

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

### Download Face2Face Masks

Source: https://github.com/ondyari/faceforensics/blob/master/dataset/README.md

Download the corresponding binary masks for the Face2Face manipulation method, indicating manipulated pixels.

```python
python download-FaceForensics.py <output path> -d Face2Face -t masks
```

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

### Generate Video from Frames with FFmpeg

Source: https://github.com/ondyari/faceforensics/blob/master/dataset/DeepFakes/faceswap-master/USAGE.md

Stitch individual image frames back into a single video file using FFmpeg. This is typically done after processing frames with Faceswap.

```bash
ffmpeg -i video-frame-%0d.png -c:v libx264 -vf "fps=25,format=yuv420p" out.mp4
```

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

### FaceForensics Dataset Folder Structure

Source: https://github.com/ondyari/faceforensics/blob/master/dataset/README.md

This outlines the directory structure for the FaceForensics++ and DeepFakeDetection datasets, indicating where original, manipulated, and downloaded videos/images are stored.

```shell
FaceForensics++ dataset
|-- downloaded_videos
    < contains all original downloaded videos, video information files and their extracted sequences
      which can be used to extract the original sequences used in the dataset >
|-- original_sequences
    |-- youtube
        < c0/raw original sequence images/videos of the FaceForensics++ dataset >
        < c23/hq original sequence images/videos >
        < c40/lq original sequence images/videos >
    |-- actors
        < images/videos from the DeepFakeDetection dataset >
|-- manipulated_sequences
    |-- Deepfakes
        < images/videos of all three compression degrees as well as models and masks after poisson image editing>
    |-- DeepFakeDetection
        < images/videos ... as well as masks >
    |-- Face2Face
        < images/videos ... as well as masks >
    |-- FaceSwap
        < images/videos ... as well as masks >
    |-- NeuralTextures
        < images/videos ... well as masks >
```

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

### Compress Data for Paper

Source: https://github.com/ondyari/faceforensics/blob/master/dataset/README.md

Compress downloaded/extracted raw images in the same manner as described in the paper. This script also includes wrapper scripts for various compression tasks.

```python
python compress.py
    -i <path to FaceForensics++ folder including original and manipulated sequences folders>
```

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

### Download Original YouTube Videos

Source: https://github.com/ondyari/faceforensics/blob/master/dataset/README.md

Download the original videos as sourced from YouTube. The downloaded file includes all videos in their original length and a JSON file with extracted frame information.

```python
python download-Faceforensics.py <output path> -d original_youtube_videos
```

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

### Extract Video Frames with FFmpeg

Source: https://github.com/ondyari/faceforensics/blob/master/dataset/DeepFakes/faceswap-master/USAGE.md

Split a video into individual frames using FFmpeg. This is useful for preparing video data for face swapping.

```bash
ffmpeg -i /path/to/my/video.mp4 /path/to/output/video-frame-%d.png
```

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

### Download Original YouTube Video Info

Source: https://github.com/ondyari/faceforensics/blob/master/dataset/README.md

Obtain only the frame locations and video information for original YouTube videos, useful for downloading them independently.

```python
python download-Faceforensics.py <output path> -d original_youtube_videos_info
```

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

### Citation for FaceForensics++ Dataset

Source: https://github.com/ondyari/faceforensics/blob/master/README.md

Use this citation when referencing the FaceForensics++ dataset in academic work. It includes author, title, booktitle, and year.

```bibtex
@inproceedings{roessler2019faceforensicspp,
	author = {Andreas R"ossler and Davide Cozzolino and Luisa Verdoliva and Christian Riess and Justus Thies and Matthias Nie{\ss}ner},
	title = {Face{F}orensics++: Learning to Detect Manipulated Facial Images},
	booktitle= {International Conference on Computer Vision (ICCV)},
	year = {2019}
}

```

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

### Create Manipulated Images from Models

Source: https://github.com/ondyari/faceforensics/blob/master/dataset/DeepFakes/README.md

Generates manipulated image sequences from trained DeepFakes models. This command can be used to recreate manipulated sequences from provided models or your own trained models. A filelist JSON can be provided to define input pairs.

```shell
python deepfakes.py
    -m create_from_models
    -i <path to models folder>
    -i2 <input folder, e.g., path to 'original_sequences' folder>
    -o <output_path>
    --python_path <path to python file, e.g., /home/<user>/anaconda3/bin/python>
    (--filelist <path to filelist json that defines the pairs from the input folder (check out our filelist splits as an example)
```

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

### Extract Compressed Video Frames

Source: https://github.com/ondyari/faceforensics/blob/master/dataset/README.md

Extract image frames from compressed videos using either ffmpeg or opencv. Use 'all' for all datasets or specify a single dataset like 'Face2Face' with 'c0' for lossless compression.

```python
python extracted_compressed_videos.py <output path> -d <"all" or single dataset via "Face2Face" or "original"> -c c0
```

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