3D Speaker Highlighting in Digital Messaging via 2D Synthesis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current messaging systems cannot effectively provide a 3D representation of speakers in video and picture messages due to limitations in recording methods and equipment, resulting in a loss of quality and inability to highlight the speaker in complex backgrounds.
Innovation Solution
A method that processes 2D images from multiple perspectives to create a 3D representation of the speaker by adjusting brightness and color, using face recognition to identify and modify sender-relevant image areas, and generating a 3D model from stored reference images or speech data, allowing for the creation of a digital graphic model that animates the mouth area.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If multiple cameras are used to capture 3D images, then the 3D quality is improved, but the device complexity and base width increase
Solution Approach 1:
The patent transforms 2D images into 3D representations by processing multiple 2D images from different perspectives and synthesizing them into a 3D model. This allows the system to achieve 3D quality without requiring multiple physical cameras in 3D space, thereby avoiding increased device complexity and base width requirements
Solution Approach 2:
The patent creates virtual copies of 2D images from different perspectives and combines them to generate 3D representations. By copying and processing 2D image data rather than capturing direct 3D data with multiple cameras, the system achieves 3D quality while maintaining simple device architecture
2Device complexity
If 2D images are used for messaging, then the device compactness is maintained, but the ability to represent speakers in 3D and highlight them in complex backgrounds is limited
Solution Approach 1:
The patent enhances 2D images by generating 3D representations of speakers within the 2D message format. This allows the system to maintain device compactness while adding 3D visualization capability that improves speaker identification and highlighting in complex backgrounds
Solution Approach 2:
The patent applies 3D representation specifically to the speaker region within the 2D image, while the rest of the image remains in 2D format. This localized enhancement maintains overall device compactness while providing improved speaker identification capability where needed
Data Source
Figure 1
Figure 2a~2b
Figure 3a~3b
AI summary
The method involves identifying a set of picture regions including one of the faces and persons in digital data of a set of messages on a digital messaging system having same sender address. A set of sender-relevant picture regions in the identified picture regions representing a sender of one of the messages is determined based on frequency of occurrence of the identified picture regions in the messages having the same sender address. The digital data of the messages is modified to emphasize one of the sender-relevant picture regions. Independent claims are also included for the following: (1) a digital messaging system comprising an identifying unit (2) a method for controlling a digital messaging system.