Invezz

Amazon joins the DeepFake Detection Challenge (DFDC) by Facebook & Microsoft

Amazon joins the DeepFake Detection Challenge (DFDC) by Facebook & Microsoft
Michael Harris
Oct 23, 2019, 04:35 AM
  • Amazon pledges $1 million to DeepFake Detection Challenge (DFDC) by Microsoft & Facebook.
  • Experts from Amazon Machine Learning solutions lab to help the researchers devise detection tools.
  • Google, Twitter, MIT, Cornell Tech, and the University of Oxford among the other noticeable supporters.

Amazon has made a recent announcement that it would be partnering with the tech giants like Microsoft and Facebook in their battle against Deepfakes. The manipulated content via such nefarious techniques has been bad news for businesses with major companies starting to focus on an effective counter-strategy.

Amazon Has Pledged $1 Million Investment To Fund Research On Potential Solutions For DeepFakes

As the newest member of the DeepFake Detection Challenge (DFDC) that was launched in September by Microsoft and Facebook in conjunction, Amazon has pledged an investment of as much as $1 million to be provided to the researchers to help support their efforts in counteracting Deepfakes. Amazon Web Services have also joined hands with DFDC partners to scrutinize the possibility of applying the Amazon S3 scalable infrastructure to host intricate datasets on cloud services that can help in the detection of Deepfakes.

According to the announcement, the researchers have been invited to submit their proposals for a grant that is expected to start at $1000 up to a maximum of $10,000 in valuation. Taking a step further, the tech giant has further added that it is open to sharing human resources as well. The Amazon Machine Learning solutions lab has professionals with years of experience who will be helping the researchers in their venture.

Other notable academic institutions that have partnered in the combat against DeepFakes include MIT, Cornell Tech, and the University of Oxford. The members so far have pledged around $10 million investment for the project. In order to help the researchers devise state of the art detection tools, a huge dataset comprising of 5000 videos was also released. The complete dataset, however, that is anticipated to be much larger, will be declared by the end of the year 2019.

Google And Twitter Are Also In Support Of Developing Innovative Methods Against DeepFakes

Owing to the threat that DeepFakes pose to the digital world, Google has also been active in helping the researchers develop detection tools. Only recently, it had released a significantly large dataset of its own that contained DeepFake videos. The officials from Google were reported stating that the dataset is aimed at developing innovative methods that can seamlessly, effectively, and efficiently filter the fake content on the internet.

Twitter has also announced on Wednesday morning, October 23rd, 2019 that it is working diligently to revise its policy regarding DeepFakes and is taking the feedback from its user base of millions of people across the world.