Odnoklassniki launched a deepfake service for personal New Year's greetings
Image: ok.ru
The social network Odnoklassniki (OK) now has a service for creating personal animation “New Image” based on deepfake technology and neural networks, with the help of which users will be able to wish each other a Happy New Year in a festive manner. This was reported in a press release received by the editorial board of “Lenta.ru”.
To create a personalized animation, users need to go to the widget, upload a photo from their computer, and select the animation they like from the list. In a few minutes, neural networks integrate the user's face into the animation and show the result in the form of a ready-made GIF animation: it can be immediately published in the news feed. At the same time, OK will not keep the original uploaded photos: they will be deleted immediately after exiting the gadget.
The launch of the New Images service was the first step in the development of OK's own deepfake technologies, the message says. Every day on the social network, tens of millions of users communicate with friends and family and share interesting content with them. With the help of technology, OK plans to improve user experience, help increase the quality and variety of published content.
Over the next year, OK plans to expand the capabilities of the service, as well as integrate it into various mechanics and products of the social network: stickers, profile photos, frames, personal congratulatory videos and others. With the help of technology, it will be possible to personalize communication with friends, as well as to try on a new style of clothing, hairstyle or accessories.
Now the service “New Image” is available in the desktop version of the social network. In the future, the function will also appear in OK mobile applications on Android and iOS.
The direction of deepfake-technologies will be one of the key ones in the work of OK with tools for intelligent processing of user-generated multimedia content (video, photo, animation). In 2022, the social network plans to launch a number of services based on computer vision, machine learning and neural networks for editing and working with the quality of user and group content.
