How to install DeepFaceLive – python application on Windows?

In this DeepFaceLive tutorial, we will guide you through the step-by-step process of installing DeepFaceLive on a Windows computer. DeepFaceLive is a great tool that can be used for real-time face swaps during PC streaming or video calls.

In the video, we have used Anaconda to create a Python environment. We tried using Python versions 3.10 and 3.11, but the application didn’t run. Therefore, please install Python 3.9 if you have a higher version of Python. After installing Python, simply activate it and run the DeepFaceLive application. If you encounter any missing modules or errors, you will need to install them manually. You may have to install PyQt6, numpy, CV2, h5py, onnxruntime, and onnx to run DeepFaceLive smoothly.

After installing all the missing modules, DeepFaceLive should run smoothly on your Windows computer. For detailed instructions and examples, please check our GitHub page.

Please try this Python application and show your creativity. At the same time, be responsible for your actions when using this application. Thank you.

STEPS:

Installing Anaconda
1. Download and Install anaconda (https://www.anaconda.com/)
2. Open Anaconda Prompt

Installing Python and running the application
1. conda create –name deepfake python=3.9 (Creates an environment named deepfake with python 3.9)

2. conda activate deepfake (activates deepfake (base) in front will change to deepfake)

3. Run DeepFaceLive
python main.py run DeepFaceLive –userdata-dir C:\Users\…..

4. Install PyQt6
pip install PyQt6 –upgrade –force-reinstall

5. Install numpy
pip install numpy

6. Install CV2
conda install -c conda-forge opencv
OR
pip install opencv-python

7. Install numexpr
pip install numexpr

8. Install h5py
conda install h5py

9. Install onnxruntime
pip install onnxruntime

10. Install onnx
pip install onnx






SUBSCRIBE

SUBSCRIBE

Welcome to Hey, Let's Learn Something!

Please sign up here to receive the latest updates on our blogs, tutorials and download section :)

Thank you for subscribing :)