Researchers specializing in artificial intelligence have virtually brought to life the subject of Leonardo da Vinci's famous Mona Lisa painting.
This was exhibited in a video that was created with the help of AI in which her head, eyes and mouth were seen to be moved by the model in the portrait. The AI researchers used only one single photo to achieve this.
Samsung's AI research laboratory in Moscow was the brains behind this achievement and the researchers made use of the latest iteration of the so-called deepfake technology of AI.
However there are concerns among some that the development and popularizing of the convincing deepfake technology could be used for nefarious purposes and be misused.
Researchers at the Samsung's institute made use of a special algorithm that was trained on a public database of 7,000 images comprising of celebrities collected from YouTube. The AI algorithm managed to bring to life a photo after mapping the facial features and movements of the celebrities which were later used on a photo. Videos of Salvador Dali, Albert Einstein, Fyodor Dostoyevsky and Marilyn Monroe were also made by the AI system.
The creations of its team of researchers were described as being "realistic neural talking heads" by Samsung in a paper outlining its work.
With respect to the video demonstrating how the system worked. A number of experts and analysts presented mixed reactions. One of them described the video as like "watching the future of SkyNet unfold".
A similar system was displayed by researchers at Tel Aviv University in 2017. Dr Supasorn Suwajanakorn had created a fake video of President Barack Obama in 2017.
While acknowledging the chances of this rising AI technology being misused, the researcher in an interview to BBC also added in the same breath that this new technology as one that can bring in good for society such as by allowing grieving relatives to create avatars of family members after they have passed away.
On earlier occasions, referring to this AI based technology, experts have said that if this technology is used to create fake videos of politicians, entire populations could be fooled and its misuse would potentially not stop at political manipulation.
Using photos of celebrities to create pornography have also been possible earlier using this deepfake technology.
"The rise of convincing deepfakes is something that could be really problematic unless we have this conversation. Members of the public need to know how easy it is to create convincing fake videos," said Dave Coplin, chief executive of AI consultancy The Envisioners,
(Source:www.bbc.com)
This was exhibited in a video that was created with the help of AI in which her head, eyes and mouth were seen to be moved by the model in the portrait. The AI researchers used only one single photo to achieve this.
Samsung's AI research laboratory in Moscow was the brains behind this achievement and the researchers made use of the latest iteration of the so-called deepfake technology of AI.
However there are concerns among some that the development and popularizing of the convincing deepfake technology could be used for nefarious purposes and be misused.
Researchers at the Samsung's institute made use of a special algorithm that was trained on a public database of 7,000 images comprising of celebrities collected from YouTube. The AI algorithm managed to bring to life a photo after mapping the facial features and movements of the celebrities which were later used on a photo. Videos of Salvador Dali, Albert Einstein, Fyodor Dostoyevsky and Marilyn Monroe were also made by the AI system.
The creations of its team of researchers were described as being "realistic neural talking heads" by Samsung in a paper outlining its work.
With respect to the video demonstrating how the system worked. A number of experts and analysts presented mixed reactions. One of them described the video as like "watching the future of SkyNet unfold".
A similar system was displayed by researchers at Tel Aviv University in 2017. Dr Supasorn Suwajanakorn had created a fake video of President Barack Obama in 2017.
While acknowledging the chances of this rising AI technology being misused, the researcher in an interview to BBC also added in the same breath that this new technology as one that can bring in good for society such as by allowing grieving relatives to create avatars of family members after they have passed away.
On earlier occasions, referring to this AI based technology, experts have said that if this technology is used to create fake videos of politicians, entire populations could be fooled and its misuse would potentially not stop at political manipulation.
Using photos of celebrities to create pornography have also been possible earlier using this deepfake technology.
"The rise of convincing deepfakes is something that could be really problematic unless we have this conversation. Members of the public need to know how easy it is to create convincing fake videos," said Dave Coplin, chief executive of AI consultancy The Envisioners,
(Source:www.bbc.com)
