Truth or Fiction? The Art of Unveiling DeepFakes in Text, Audio, Image, and Video.

In a world full of deepfakes, how can you find the truth and detect whether a text, audio file, image, or video is genuine or not?

In this presentation, I will provide an overview of the tools that can be used to detect DeepFakes.

For text, we will introduce how to identify whether texts generated via ChatGPT are real or not using tools such as GPT Output Detector, GLMT, and GPTZero.

I will present some ocular methods to distinguish between genuine and fake images generated via DALL-E.

For audio, I will introduce methods based on ML, DNN, and POI.

Finally, I will demonstrate how frame-by-frame analysis, multimodal methods, POI, and commercial products can be used to detect deepfakes in videos.

Håkan Silfvernagel

AI Specialist, Microsoft AI MVP

Oslo, Norway


Please note that Sessionize is not responsible for the accuracy or validity of the data provided by speakers. If you suspect this profile to be fake or spam, please let us know.

Jump to top