The AI that lets you alter any face with a quick doodle: Change hairstyles, smiles and add accessories like sunglasses to images using idiot-proof software made in South Korea
- The system was developed at the Telecommunications Research Institute
- The technology lets people change photos of themselves through doodles
- By drawing an altered jawline on the image the system interprets the change
- In the study it was used on photos of celebrities including Chris Hemsworth
If you want to make changes to the way your face, hair, and clothing appear in photos you’ve posted online, help could soon be at hand.
Artificial Intelligence (AI) software under development in South Korea lets you make alterations with the flick of a finger.
Users can draw the changes they want – such as higher eyebrows, a different jawline or new earring – onto images directly.
The system interprets these doodle to make realistic changes to the photo underneath.
Scroll down for video
The new system, published in a scientific paper, allows users to change photos by doodling onto them. Seen above: Amy Adams’s eyebrow, nose and jawline is changed
Details of the new software were released in a preview of a study by Youngjoo Jo at South Korea’s Electronics and Telecommunications Research Institute, due to be published later this year.
Graphics released with the study show a woman’s picture with a fringe, make-up and larger nose drawn on.
The system interprets these doodles and changes the faces accordingly, making them look more realistic.
- Disgraced Chinese scientist who performed ‘monstrous’… Scientists who created Dolly the sheep reveal plans for… Chinese scientist who claims he made world’s first… A ‘monstrous’ experiment or a huge leap for science? Experts…
Share this article
Researchers trained the software using images of celebrities, including Amy Adams who has her eyebrow enlarged, nose shifted and jawline changed by the system.
Australian heartthrob Chris Hemsworth has his eyes coloured and mouth shifted into a smile.
The system interprets simple doodles, allowing users to change their appearance. Seen above: Julia Roberts has her eyes and lips changed by the system
Photos of Julia Roberts, actor Josh Henderson, and others have also been used, showing them with lips and hairlines changed, glasses added and removed and their eyes shifted.
It is not clear yet if the system will be made available to the public.
In the abstract, the paper reads: ‘Contrary to existing methods, our system wholly utilises free-form user input with colour and shape.
‘This allows the system to respond to the user’s sketch and colour input, using it as a guideline to generate an image.’
The full findings of the study were released on the pre-print server Arxiv.org.
Developed by Youngjoo Jo at South Korea’s Electronics and Telecommunications Research Institute, the new system offers the chance to add and change features on an image
The system will also be able to re-interpret drawings of faces and try to generate a more realistic image of the person
HOW DOES ARTIFICIAL INTELLIGENCE LEARN?
AI systems rely on artificial neural networks (ANNs), which try to simulate the way the brain works in order to learn.
ANNs can be trained to recognise patterns in information – including speech, text data, or visual images – and are the basis for a large number of the developments in AI over recent years.
Conventional AI uses input to ‘teach’ an algorithm about a particular subject by feeding it massive amounts of information.
AI systems rely on artificial neural networks (ANNs), which try to simulate the way the brain works in order to learn. ANNs can be trained to recognise patterns in information – including speech, text data, or visual images
Practical applications include Google’s language translation services, Facebook’s facial recognition software and Snapchat’s image altering live filters.
The process of inputting this data can be extremely time consuming, and is limited to one type of knowledge.
A new breed of ANNs called Adversarial Neural Networks pits the wits of two AI bots against each other, which allows them to learn from each other.
This approach is designed to speed up the process of learning, as well as refining the output created by AI systems.
Source: Read Full Article
Giant duck sparks rescue emergency as three men drift out to sea on inflatable
Kilauea Erupts in Hawaii With ‘Incandescent’ Glow
‘Killer’ great white sharks more Gums than Jaws – as they rarely attack humans
AI reincarnations of dead people to haunt and scam bereaved families online
‘Somerset gimp’ unmasked as man alleged to be rubber pervert appears in court
New contraceptive injection could help manage stray cat populations
Only those with a high IQ can tell which of these boxes are the heaviest
Teen ‘arrested over vile dog sex vid’ slams people for ‘being cruel to animals’