Invention Title:

METHOD, SYSTEM, AND MEDIUM FOR ENHANCING A 3D IMAGE DURING ELECTRONIC COMMUNICATION

Publication number:

US20250029346

Publication date:
Section:

Physics

Class:

G06T19/20

Inventors:

Applicant:

Smart overview of the Invention

The disclosed method and system focus on enhancing holographic communication by improving the representation of participants in a 3D format during electronic communication. Visual and verbal data of a participant are collected as a current sample, which is then encoded into a vector comprising multiple states. These states are updated using a generative artificial intelligence model, resulting in a predicted outcome that refines the visual data for rendering a 3D representation in real-time or future frames.

Technical Challenges and Solutions

Holographic communication faces challenges, especially in accurately capturing dynamic facial expressions and movements. Existing methods struggle with reconstructing expressive facial features crucial for effective communication. The proposed system addresses these issues by utilizing AI to predict and update missing or obscured parts of the hologram, ensuring a more complete and realistic representation of participants.

Method Details and Embodiments

The method involves encoding visual and verbal data into vectors representing different states like texture, shape, and facial expressions. These vectors are inputs to an AI model that predicts updates to create a 3D representation. The approach includes using past samples to refine predictions, employing neural networks for reconstruction tasks, and accommodating various data formats for comprehensive modeling.

System Architecture

The system comprises a network interface, processor, and storage medium with executable code for implementing the method. It may include cameras for capturing participant images and displays for showing reconstructed 3D representations. This setup supports real-time holographic communication by processing captured data through neural networks to produce photo-realistic 3D images.

Advantages Over Conventional Methods

Unlike traditional 2D conferencing or avatar-based systems, this approach aims for photo-realistic 3D communication by leveraging AI to fill in gaps left by standard capture methods. It enhances immersion and non-verbal communication, addressing limitations of prior technologies by providing realistic, dynamic representations of participants during virtual interactions.