Artificial intelligence or AI can be a useful Instrument to help take the burden off our shoulders when it comes to trivial and mundane tasks.
Companies like Google have employed AI and machine learning how to emulate individuals for simple tasks like reserving a salon appointment, with products like google Duplex. AI algorithms also have been used to animate 2D objects like talking heads of individuals, but this typically takes a huge dataset of samples to train the neural network to adapt to the objects or in this case, that the person’s behavior and mannerisms. This process just got a lot easier thanks to Samsung AI Center’s Moscow chapter, which has released a paper that shows a highly realistic talking head design can be obtained by feeding the algorithm together with as little as one sample photo.
The algorithm utilizes a technique of using”landmarks,” that is a skeletal framework of a persons face learnt by glancing through a huge dataset of videos, which is used as the backbone for its synthesised face. The precision of this synthesised face is dependent upon the number of source pictures you feed the algorithm.
The paper shows the difference between versions trained using 32 samples, eight samples, and one sample. The output from just one sample is obviously has a couple of artefacts round the face, helping you see that it’s not a genuine video but an AI-generated one. However, the fact we could even find such output with one sample is still shocking. The video and the paper indicates the algorithm successfully creating synthesised faces of celebrities and even paintings, like the Mona Lisa.
Such a technique can be utilised to bring you more realistic appearing’Memojis,’ that can be Apple’s version of an Animoji however of yourself. The creative potential of this is endless. However, such technologies may also be misused if fallen in the wrong hands. Celebrities have often been the target of these”deepfake,” images and videos, which is aimed to tarnish their social standing. It could of course, also be used to make fake news.
Earlier this year, Hollywood actress Scarlett Johansson was outspoken against the dangers such”deepfakes,” following her face had been grafted into pornographic videos. Making such bogus photos and videos is not a new dilemma, but the fact it may be done this easily now, is quite worrying. Social media platforms like Twitter, Reddit as well as adult entertainment portals such as Pornhub have begun cracking down of AI-generated”deepfakes.”