Detection of Fake Facial Images and Changes in Real Facial Images

Computer Vision techniques are widely used in the entertainment industry, helping to create more realistic effects in games and movies. They can recognise objects, characters, and player movements in video games. This allows games to react to player behaviours more intelligently, providing more dynamic and engaging experiences. Additionally, applying deep learning techniques combined with Computer Vision supports generating automatic special effects, such as adding interactive effects to live broadcasts. Unfortunately, such methods can generate, modify, and falsify information, such as swapping faces in a photo or video recording. Social media has many counterfeits and modifications of content known as fake news. The article proposes a method for detecting modified, real facial images and artificially generated facial images based on convolutional neural networks. Our technique allows for classifying facial photos into one of three classes: real faces, real faces with applied modifications (using photo editing software), and artificially generated facial images.