Detect deepfake faces in videos and pictures with an accuracy of approximately 93%. Our technology does not only classify an input as fake, but also explains the reasoning behind its decision by using Activation Maps.
Access our ready-to-use technology by using our customisable APIs to integrate DeepDetector into your current workflow.
All forms of analyses take place in accordance with European laws and regulations regarding data protection, privacy, and responsible AI.
DeepDetector is trained with the largest scientific or self-made representative datasets. This way we capture significant information input more efficiently than any other algorithm.
DeepDetector filters all kinds of FaceSwaps and other AI-manipulations by looking for traces of alterations in existing (camera-made) pictures and videos.
AI-generated visual content, such as StyleGANs are not distinguishable for the unaided human eye. DeepDetector analyses characteristics of AI-generated content to detect them.
Image analysis time
Explainable results
Deepfake types can be detected
Percentage of correct predictions (of both real and fake)
Percentage of fake classifications that were actually fake
Percentage of actual fakes that are correctly predicted
DeepDetector is a deep learning network designed and trained to recognise AI-generated or AI-manipulated faces. DeepDetector can be seen as an artificial neural network designed to spot forgeries and traces generated by computers.
It is able to do this, because the DeepDetector has seen hundreds of thousands of real and deefake images. It has trained itself on these images and has learned to find the subtle differences between a photo made with a real camera or a photo synthesised by a computer.