Home » News » New AI Model Can Make You Look Younger or Older in Photos Without Losing Your Identity

New AI Model Can Make You Look Younger or Older in Photos Without Losing Your Identity

(Image Credit Google)
(Image credit- IFL Science) Researchers from the NYU Tandon School of Engineering have created a novel artificial intelligence technique that can alter a person's apparent age in images while preserving their distinctive identifying features. This approach represents a significant improvement over existing AI models that can alter a person's appearance by making them appear younger or older without keeping their unique biometric identifiers.

New AI Model That Preserves Unique Identity Features

Sudipta Banerjee, a research assistant professor in the Computer Science and Engineering (CSE) Department, and colleagues trained a kind of generative AI model, a latent diffusion model to "know" how to carry out identity-retaining age transformation in a paper that was published on the pre-print server arXiv and will be presented at the proceedings of the IEEE International Joint Conference on Biometrics (IJCB). [caption id="" align="aligncenter" width="875"]Disney built an AI that can easily make actors look younger or older | Engadget Image credit- Engadget[/caption] This AI model has been taught to carry out age modifications while maintaining identity. In order to accomplish this, the researchers overcame a problem indigenous to this field: gathering a sizable dataset of images of people from over a decade. In contrast to this technique, the researchers used a limited dataset of images representing a human, augmented by another set of images with captions specifying the represented person's age group, such as kid, teenager, young adult, middle-aged, elderly, or old. Particularly, this collection included images of famous people taken over the course of their lives. The AI model acquired the ability to identify the distinctive biometric characteristics that set individuals apart from the initial sample.

How Does It Work?

The age-categorized images made it easier for the model to understand the connection between age and photos. As a result, by entering a desired age in a text prompt, the trained model could replicate the aging or rejuvenation process. For altering human facial images, the researchers used the "DreamBooth" method, which involves making progressive alterations using a combination of neural network elements. [caption id="" align="aligncenter" width="1600"]Disney's New AI Software Can Easily Make Actors Look Younger or Older | PetaPixel Image credit- PetaPixel[/caption] This technique uses the underlying data distribution to introduce and remove noise, or random changes, from photographs. The researchers controlled the image-generating process by adding text prompts and class labels, with an emphasis on the preservation of identity-specific information and image quality. The neural network model might be adjusted more easily due to the use of various loss functions. Through studies that generated human facial images displaying aging-related alterations and contextual differences, the researchers demonstrated the effectiveness of their approach. The researchers claimed that their approach worked better than competing methods, showing a reduction of up to 44% in the rate of inaccurate rejections throughout the recognition process. Meanwhile, the team's research was published as a pre-print on arXiv and was scheduled to be presented at the IEEE International Joint Conference on Biometrics (IJCB). Also read: Microsoft officially introduces Bing AI Chat to Google Chrome

By Raulf Hernes

If you ask me raulf means ALL ABOUT TECH!!


In the ever-changing world of technology and retai...


In a bid to capture the attention of users and dri...


Apple is preparing for a game-changing move with i...


Google has been making huge headways in artificial...


Elon Musk's artificial intelligence firm, xAI, is ...


In a digital showdown that has captured the attent...