ICrypto

Hotest Blockchain News in First Media Index

Disney's innovative AI can quickly make actors appear younger or older

Researchers at Disney have built an artificial intelligence toolthat can make it easier for an actor to appear a different age on screen.Although digital artists can still make necessary modifications to make the effectsin a scene look as realistic as possible, the artificial intelligence systemcan handle most of the aging effects.

Re-aging characters in films using AI

In movies and advertisements, it is costly to create photorealisticdigital re-aging and requires artists to go through each scene arduously,frame-by-frame to manually change the character’s appearance and likeness. Inthe past, there have been attempts to automate the strenuous, time-consuming,and expensive task, using AI, both machine learning and neural networks.

The Disney researchers said that other systems currently available“typically suffer from facial identity loss, poor resolution and unstableresults across subsequent video frames.” The research team mentioned that theirmethod offers “the first practical, fully-automatic and production-ready methodfor re-aging faces in video images.”

The face re-aging network, or FRAN

The team calls their AI system FRAN, an acronym for facere-aging network.

An actor re-aged using FRAN on the right, compared to only wearing a costume.

Production-Ready Face Re-Aging for Visual Effects/Disney Research Studios 

FRAN is a neural network that incorporates neural face modelsfrom being pre-trained on thousands of images of different faces without anyage pairing. Neuralface models are created using deep neural networks and are used in digitalwork and facial animation.

Researchers mentioned that it would be impossible to trainFRAN on datasets using real people because that would require numerous pairs ofimages showing people with the same facial expressions, on the same backgroundsat two known ages.  FRAN was trainedusing datasets from thousands of randomly generated faces to collect the necessaryinformation for re-aging actors.

The new research conducted by Disney Research Studios shortensthe aging process in video creation. Disney Research Studios focuses on innovationsin filmmaking by using deep learning, artificial neural networks, and othersubsets of artificial intelligence to use in videos.

The neural network can analyze an actor’s headshot andpredict what part of the face would potentially be affected by aging. It thenapplies the aging properties, such as wrinkles for older effect or skinsmoothing for a younger aesthetic. The process layers the AI- generated effecton top of the original face. The ability to use AI in the images speeds up the processand efficiently creates realistic images.

Most Popular

Re-aging actors in 2D images

The novel AI model FRAN is able to generate images byimproving them on a 2D re-aging workflow, according to researchers at Disney. “FRANcurrently only improves upon the predominant 2D re-aging workflow, whereas inparticular cases a 3D re-aging solution may be preferable for more elaboratelevels of control including, for instance, relighting and other physicallybased manipulations,” the researchers said in the study.

Any images thatrequire enhancements in a 3D format are still being explored. The team hopesthat they will one day be able to also use artificial intelligence on 3Dre-aging as well. “We believe that these limitations also represent excitingopportunities for additional improvements in future work.”

Challenges using a U-NET architecture

Researchers mentioned a few limitations with the current AItool. There are a few challenges, such as generating large changes on images. Theteam notes the complexities are often difficult in typical U-Net architectures,a form of deep learning. A U-Net architecture is a symmetric architecture usedin image segmentation, created in 2015, specifically at the time to solveissues within biomedical image segmentation.It is now used in various artificial intelligence systems. It is named for theshape of its network, forming a U-shape and consisting of the usual AI elementsof training the network using input datasets to produce output.

A U-Net architecture from the original 2015 study

U-Net: Convolutional Networks for Biomedical Image Segmentation/Lecture Notes in Computer Science  

FRAN creates the images using a U-NET architecture incorporating alarge number of datasets. The AI system can create realistic and re-aged imageswithin a range of 18 and 85 years old.

Due to the limitation of creating large changes, stated bythe researchers, there is difficulty in aging characters from extremely youngages to the present age of the actor. They recommend using other methods, such as LATS or DLFS,known as deep learning from scratch in AI, in these cases.

The research team also noted that another limitation when agingcharacters using AI, is the difficulty of graying scalp hair on characters, aneffect that the AI system currently is unable to capture from the datasets. Anadditional limitation is re-aging characters while including possible changesin Body Mass Index (BMI) as they age. “Similarly, re-aging can also introducevariation in body mass index (BMI) whose effects on the face we currentlycannot control,” the team stated.

The future of FRAN

Despite the challenges, the team is optimistic about usingthe AI to improve the re-aging process in films and create a simpler, quicker processthey can continue to build upon through future advancements. “FRAN has thepotential to improve existing re-aging workflows,” the researchers said, “reducingthe time it takes to re-age complete shots from a matter of days to just a fewhours or even minutes, facilitating the creation of high-quality visual effectsat scale.”

SHOW COMMENT (1)
For You
culture
Digital avatars will allow 8 billion people to be a part of COP27 and demand action

Chris Long is no stranger to getting millions involved in social causes and now want to leverage technology to involve billions of people.

Ameya Paleja | 11/15/2022
scienceA brief guide to looking for aliens
Grant Currin| 7/31/2022
innovationThis space debris removal company is on a singular mission to clean Earth's orbit
Deena Theresa| 9/14/2022
More Stories
health

World’s first screening test for pancreatic cancer sees worms sniff out tumors

Loukia Papadopoulos| 12/4/2022
science

Breakthroughs in genomics show that 'junk' DNA is incredibly important

Grant Currin| 8/4/2022
innovation

Backyard Scientist’s rocket-knife hits the target at 400mph

Loukia Papadopoulos| 12/3/2022
Share
 05.12.2022

Hotest Cryptocurrency News

End of content

No more pages to load

Next page