Invention Title:

SYNTHESIZED REALISTIC METAHUMAN SHORT-FORM VIDEO

Publication number:

US20240292069

Publication date:
Section:

Electricity

Class:

H04N21/816

Inventors:

Assignee:

Applicant:

Drawings (4 of 9)

Smart overview of the Invention

Techniques for creating synthesized realistic metahuman short-form videos leverage photorealistic representations of individuals sourced from various media, including videos and photographs. By analyzing viewer data such as purchase and viewing history, the system selects an appropriate individual to feature in the video. Machine learning is then employed to isolate the individual's representation, which is used to develop a three-dimensional (3D) model through a game engine.

Creation of Realistic Synthetic Performances

The process involves combining the 3D model with animations generated by the game engine to create a synthetic performance. This performance can also incorporate the voice of the selected individual, enhancing realism. The final product is rendered as a short-form video that can be integrated into a metaverse environment, providing an interactive experience for viewers.

Importance of Short-Form Videos in Modern Communication

Short-form videos serve as vital communication tools across various sectors including advertising, education, and business. As audiences become more discerning, the choice of spokesperson significantly influences marketing success. Influencers and trusted sources play crucial roles in shaping consumer decisions, making the strategic selection of hosts essential for effective engagement.

Streamlining Video Production with AI

Traditionally, producing short-form videos involves extensive recording and editing processes. However, AI and machine learning technologies streamline this by enabling real-time interactions with 3D models of hosts. This capability allows for dynamic presentations during livestream events, enhancing viewer engagement while also creating content that can be reused across different platforms.

Technical Process and Features

The technical methodology includes accessing photorealistic representations, isolating them through machine learning, and creating detailed 3D models. The isolated images capture intricate details such as facial features and body proportions. The synthesized performances can be rendered into short-form videos that feature e-commerce elements, allowing viewers to make purchases directly from the content they engage with.