With the new feature, if experts are unsure of the accuracy or cannot confirm the accuracy of the data, TikTok will flag such videos as containing “unverified content”
Author: DFC
A way to combat fakes for Clubhouse could be to tag the verified account and enable application-level identification that protects against voice forgery using voice biometrics technologies. To obtain such a label, the social network can oblige the user to leave a “voice cast”, and at the level of the application itself create voice biometrics
A white paper published by cybersecurity firm Nisos, sketches five incidents involving deepfake audio attacks
There is also a positive side to the difficulty in recognizing masked people. The problem faced by authentication services is to be solved by facial recognition software
Of course, we sent out a link to Product Hunt to everyone we know. I asked the guys on Mesto social network and Silicon Pravda Telegram chat to support us. But most of the voices came from strangers, which was very encouraging
Editing individual elements in a video requires frame-by-frame changes. This is usually done manually, like in animation, or using special tools that have limited functionality and can only perform certain programmed actions. A development team at Google and Oxford University was able to split each video frame into separate layers and taught AI to recognize people or other moving objects in them
Thanks to the development of artificial intelligence, it is becoming more difficult to understand where is reality and where is fake. If last year deepfake technology caused concern, now there is a fake face phenomenon, that is, the creation of images of non-existent people indistinguishable from real images. According to experts, this technology is actively used for political propaganda
According to the terms of the tender, the winner will have to carry out work on the topic “Research of possible ways to identify signs of intra-frame video editing, performed using neural networks.” In addition, the contractor will have to draw up terms of reference for the development of a hardware and software complex for analyzing video fakes
The developers claim that the accuracy of the instrument can reach 99.9%, but the detector is not in all cases “sure” almost 100% in its decision. Also in the future, it is planned to train the system to distinguish whether other objects have changed in the video or photo
“Don’t talk to strangers,” our parents advised us. In an era of double life: real and virtual, someone must also tell us, “Don’t share everything on Facebook!”