To solve the problem, Facebook announced Thursday with Partnership on AI,Microsoft 公司总部所在地：美国 主要业务：软件 Deepface Test Challenge (DFDC) was launched in collaboration with Cornell Institute of Technology, Massachusetts Institute of Technology, Oxford University, University of California, Berkeley, University of Maryland Parker and New York State University Albany. The goal of this challenge is to create a physical data set that can be used for training models that can accurately detect and report Deepfake.
The social networking giant will donate more than $10 billion to fund the job, and hopes to inspire researchers and organizations to come up with new ideas to counter false information on social media and the Internet. Mike Schroepfer, Facebook's chief technology officer, wrote:
We hope to promote more research and development in this area, and ensure that there are better open source tools to detect Deepfake.. Deepfake Detection Challenge will include a data set and rankings as well as incentives to encourage the industry to create new ways to detect and prevent media manipulated by AI from being used to mislead others.
Researchers at these institutions will measure the effectiveness of datasets and models trained with them at the International Conference on computer Vision (ICCV). The resulting complete dataset will be published at the Neural Information processing system meeting (NeurIPS) in December.
In addition, Facebook clarified that it would not use any user data to create the dataset and that only agreed and paid participants could participate in the challenge. For more details about the official announcement, you can clickhereRead blog articles. For more information about challenges, visitDeepface Detection Challenge Website.