DevDays Moscow 2022 

Martin Förtsch

Должность: Principal Consultant

Компания: TNG Technology Consulting

Страна: Germany

Биография

Martin Förtsch is an IT consultant of TNG Technology Consulting GmbH based in Unterföhring near Munich who studied computer sciences. Workwise his focus areas are Agile Development (mainly) in Java, Search Engine Technologies, Information Retrieval, and Databases. As an Intel Software Innovator and Intel Black Belt Software Developer he is strongly involved in the development of open-source software for gesture control with 3D-cameras like e.g. Intel RealSense and has built an Augmented Reality wearable prototype device with his team based on this technology. Furthermore, he gives many talks at national and international conferences about Artificial Intelligence, the Internet of Things, 3D-camera technologies, Augmented Reality, and Test Driven Development as well. He was awarded the Oracle JavaOne Rockstar.

Доклад

Pushing Deepfakes to The Limit — Fake Video Calls With AI

Today’s real-time Deepfake technology makes it possible to create indistinguishable doppelgängers of a person and let them participate in video calls. Since 2019, the TNG Innovation Hacking Team has intensively researched and continuously developed the AI around real-time Deepfakes. The final result and the individual steps towards photorealism will be presented in this talk.

Since its first appearance in 2017, Deepfakes have evolved enormously from an AI gimmick to a powerful tool. Meanwhile, different media outlets such as «Leschs Kosmos», Galileo, and other television formats have been using TNG Deepfakes.

In this talk we will show the different evolutionary steps of the Deepfake technology, starting with the first Deepfakes and ending with real-time Deepfakes of the entire head in high resolution. Several live demos will shed light on individual components of the software. In particular, we focus on various new technologies to improve Deepfake generation, such as Tensorflow 2 and MediaPipe, and the differences in comparison to our previous implementations.

Ключевые слова

🔑 Deepfakes
🔑 OpenCV
🔑 Transfer Learning
🔑 Generative Adversarial Networks

« Назад