Recently, a research team from the University of North Carolina and the University of Maryland introduced a new technology called "MyTimeMachine" (MyTM), an advanced system capable of personalized facial age transformation. By simply uploading 50 photos, the system can learn your facial features and estimate how you would look at different ages.
This technology combines global aging models with users' personal selfies, allowing for high-fidelity age reversal (de-aging) and aging effects while maintaining individual identity characteristics.
As technology advances, facial aging has become a widely discussed topic. Traditional facial age transformation methods often fail to accurately reflect an individual's appearance at the target age, resulting in generated images that can be far from reality. The innovation of MyTimeMachine lies in its ability to train an adapter network using 50 selfies provided by the user, enabling personalized age transformation.
The system works by first using a global age encoder to generate basic facial features, and then combining these features with the user's personal photos through the adapter network to create new facial images. The research team has also designed three loss functions to enhance the personalization capabilities of the adapter network, ensuring that the generated images maintain the user's identity characteristics throughout the aging process. This technology is not only applicable to static images but can also be extended to video processing, achieving high-quality, identity-preserving, and temporally consistent aging or de-aging effects.
In practical applications, this personalized age transformation technology can be widely used in various fields such as film and television effects, advertising, and more, achieving a more realistic visual presentation. Users can easily experience the age transformation process by uploading their personal photo sets, enjoying the magical changes brought by technology.
Project link: https://mytimemachine.github.io/
Key Points:
🌟 MyTimeMachine combines global aging models with users' personal selfies to achieve personalized age transformation.
📸 Only 50 personal photos are needed to train the adapter network, generating realistic age change effects.
🎥 This technology is applicable not only to static images but can also be extended to video, achieving high-quality, identity-preserving aging effects.
