Inside ObEN Labs - Tracking Facial Movements with AI

Every day, ObEN's world-class team of researchers and engineers work to develop cutting-edge Personal AI (PAI) technology in the areas of speech, computer vision, and natural language processing.

In the latest inside look at our technology, Computer Vision Research Scientist Dr. Peng guides us through how ObEN uses AI to track and map your facial movements from a camera to your own PAI, creating uniquely expressive PAIs that can move and react in real-time.

Every advancement from our research teams goes toward making your PAI smarter, more capable, and more useful to you. Learn more about ObEN's proprietary AI technology and our adoption of the Project PAI blockchain to secure and authenticate our Personal AI in a new digital economy at

Some teams have tried to use AI to combat Deepfake. The convolutional neural network (CNN) is used to determine whether the video has been processed. The accuracy rate is between 84% and 99%. Because most of the current Deepfake videos have glitches and noise. This feature makes their current AI accuracy rate very high, but for those who really want to use fake and real Deepfake movies, they often collect a huge amount of data and can produce high-resolution, high-quality Deepfake movies. In this case, This method does not seem to work.

I would like to know how PAI guarantees the security of data.