Deep-Stream-Cam

**Deep-Stream-Cam** enables seamless real-time face swapping and video deepfakes with just a single image and a single click. Leverage AI to effortlessly transform video content in an instant and explore advanced visual effects.

Disclaimer

This software aims to enhance the AI-generated media industry, supporting tasks like animating custom characters or creating digital models.

Important Notes:

• The software has built-in checks to prevent processing inappropriate media (e.g., nudity, war footage).
• Users must respect ethical guidelines and obtain consent when using real people's faces.
• The software may be modified or shut down if legally required.

Quick Start - Pre-built Versions

For Windows / NVIDIA GPU:

• Download Pre-built Version (Windows)

For Mac / Apple Silicon:

• Download Pre-built Version (Mac)

These pre-built versions are ideal for non-technical users who want a quick setup. Manual installation is also available for advanced users.

TLDR: Live Deepfake in 3 Easy Steps

1. Select a face: Choose the face to swap.
2. Choose your camera: Select the camera source.
3. Press Live!: See the magic in real-time!

Features & Uses - Real-time Face Swapping

1. Mouth Mask

• Retain original mouth movements for natural realism.
• 

2. Face Mapping

• Swap faces across multiple subjects in a scene.
• 

3. Movie Mode

• Replace faces in movies in real-time.
• 

4. Live Shows

• Perfect for live performances or streaming.
• 

5. Memes

• Create viral memes with ease.
• 
• Created using Deep-Stream-Cam

Installation (Manual Setup)

Note: Installation requires technical skills. For easier use, consider the pre-built versions.

Click for Installation Instructions

Requirements

• Python 3.10 (recommended)
• pip
• git
• ffmpeg
• Visual Studio 2022 Runtimes (Windows)

Steps to Install

1. Clone the Repository

git clone https://github.com/hacksider/Deep-Live-Cam.git
cd Deep-Live-Cam

2. Download Models

   • GFPGANv1.4
   • inswapper_128_fp16.onnx

   Place these files in the models folder.

3. Install Dependencies

pip install -r requirements.txt

4. Run the Application

python run.py

GPU Acceleration (Optional)

For enhanced performance, use GPU acceleration for processing:

CUDA (NVIDIA GPU)

1. Install CUDA Toolkit.
2. Install dependencies:

pip uninstall onnxruntime onnxruntime-gpu
pip install onnxruntime-gpu==1.16.3

3. Run the application:

python run.py --execution-provider cuda

CoreML (Apple Silicon)

1. Install dependencies:

pip uninstall onnxruntime onnxruntime-silicon
pip install onnxruntime-silicon==1.13.1

2. Run:

python run.py --execution-provider coreml

Usage

Image/Video Mode

1. Run python run.py.
2. Select source and target images/videos.
3. Click "Start" to process.

Webcam Mode

1. Run python run.py.
2. Choose source face image.
3. Click "Live" to start webcam mode and preview.
4. Stream with OBS or any screen capture tool.

Command-Line Arguments

Here’s a quick overview of the available command-line options:

• -s SOURCE_PATH: Select a source image.
• -t TARGET_PATH: Choose a target image or video.
• --mouth-mask: Use the mouth mask feature.
• --live-mirror: Live camera mirror.
• -v, --version: Show version info.

For full list, refer to the CLI documentation.

Press Coverage

Deep-Stream-Cam has garnered attention for its groundbreaking AI-powered face-swapping capabilities:

• Ars Technica: "Deep-Live-Cam goes viral"
• PetaPixel: "Deepfake AI Tool Lets You Become Anyone in a Video Call"
• NewsBytes: "This free AI tool lets you become anyone during video-calls"

Credits

Thanks to the following contributors:

• ffmpeg: for video processing.
• deepinsight: for face detection models.
• havok2-htwo: for webcam integration.
• Special thanks to all contributors for their dedication and support.

---