18149007. GENERATING VIRTUAL HAIRSTYLE USING LATENT SPACE PROJECTORS simplified abstract (Snap Inc.)
Contents
GENERATING VIRTUAL HAIRSTYLE USING LATENT SPACE PROJECTORS
Organization Name
Inventor(s)
Aleksandr Belskikh of London (GB)
Menglei Chai of Los Angeles CA (US)
Antoine Chassang of London (GB)
Pavel Savchenkov of London (GB)
GENERATING VIRTUAL HAIRSTYLE USING LATENT SPACE PROJECTORS - A simplified explanation of the abstract
This abstract first appeared for US patent application 18149007 titled 'GENERATING VIRTUAL HAIRSTYLE USING LATENT SPACE PROJECTORS
Simplified Explanation: The technology described in the patent application uses a Generative Adversarial Network (GAN) model to create a realistic image of a face with 3D virtual hair. It then blends this image with a second image containing realistic virtual hair to generate a new image with lifelike hair.
- The technology generates a first image of a face using a GAN model.
- It applies 3D virtual hair to the first image to create a second image with 3D virtual hair.
- The second image is projected into a GAN latent space to generate a third image with realistic virtual hair.
- A blend of the realistic virtual hair with the first image is performed to create a new image with new realistic hair.
- A neural network is trained to receive the second image with 3D virtual hair and provide an output image with realistic virtual hair.
- The trained neural network can generate a specific output image with realistic hair based on a particular input image with 3D virtual hair.
Potential Applications: This technology could be used in virtual reality applications, gaming, digital art, and fashion design to create realistic virtual hair on digital avatars or characters.
Problems Solved: This technology addresses the challenge of creating realistic virtual hair on digital images, enhancing the visual quality and realism of virtual characters.
Benefits: - Improved realism in digital images - Enhanced visual quality of virtual characters - Increased creativity in digital art and design
Commercial Applications: Title: Advanced Virtual Hair Generation Technology for Digital Applications This technology could be utilized in industries such as gaming, virtual reality, animation, and fashion design to enhance the visual appeal and realism of digital content. It could also be integrated into beauty and hairstyle simulation apps for virtual try-on experiences.
Prior Art: Prior research in the field of computer graphics and image processing may have explored similar techniques for generating realistic virtual hair on digital images. Researchers and developers in the field of GANs and neural networks may have also worked on related technologies.
Frequently Updated Research: Researchers in the field of computer graphics and artificial intelligence are continuously exploring new methods and algorithms for improving the realism and quality of virtual hair generation in digital images. Stay updated on the latest advancements in GAN models and neural networks for virtual hair rendering.
Questions about Virtual Hair Generation Technology: 1. How does this technology improve the visual quality of digital images compared to traditional methods? 2. What are the potential limitations or challenges faced in implementing this technology in real-world applications?
Original Abstract Submitted
The subject technology generates a first image of a face using a GAN model. The subject technology applies 3D virtual hair on the first image to generate a second image with 3D virtual hair. The subject technology projects the second image with 3D virtual hair into a GAN latent space to generate a third image with realistic virtual hair. The subject technology performs a blend of the realistic virtual hair with the first image of the face to generate a new image with new realistic hair that corresponds to the 3D virtual hair. The subject technology trains a neural network that receives the second image with the 3D virtual hair and provides an output image with realistic virtual hair. The subject technology generates using the trained neural network, a particular output image with realistic hair based on a particular input image with 3D virtual hair.
