Google releases FaceForensics dataset to help researchers fight deepfake technology

Many people are worried that the new Deepfake technology will make it difficult to distinguish between fake and real audio, video, and image. Now, Google wants to make it easier for people to know whether these things are real or fake, and to contribute to the detection of audio, video and image authenticity. Google said that although many fake videos and images are originally humorous, they may harm individuals and society.

Google believes that these problems are very serious, and has released a synthetic voice data set to support the development of related software to detect the authenticity of the audio. Google said that more than 150 research institutions and industry organizations downloaded the data set. The same data set is now available to the public.

This week, Google and Jigsaw launched a new visual deep false data set. The new data set has been integrated into the relevant benchmark software, which will help identify the fake video. Users can now download this dataset on the FaceForensics Github page.

Google said that the field is progressing rapidly and that it will be added to the data set as DeepFake technology evolves. Google also promises to continue to work with partners in this field. Google believes that supporting a thriving research community will mitigate the potential harm caused by the abuse of DeepFake technology.