AdVerif.ai is joining forces with leading Machine Vision researchers to compete in the Deepfake Detection Challenge organized by Facebook and Microsoft. Our team has already developed an initial prototype based on the FaceForensics dataset using the AlexNet Convolutional Neural Network.
With prizes in Millions of dollars and the opportunity to work with Facebook, one can expect that the competition will garner a lot of interest and attract great minds to address the deepfakes challenge. Perhaps it will have the same effect as the “Netflix Prize” had on recommendation systems, kick-starting a new generation of algorithmic advancements.
One of the main challenges for researchers in this field has been the lack of vast amounts of examples, needed to train deep neural networks. As part of the challenge, participants will get access to thousands of video examples, generated especially for the competition. We look forward to leveraging this new data and coming up with creative new ways of detection to achieve the best performing models.
Good luck to all the participating teams. May the best deepfake detector win!
