Multi-View Display Voice Separation for Accurate Speaker Commands

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Conventional voice recognition technologies in display devices struggle with accurately distinguishing multiple speakers and handling interference sounds, leading to repeated voice recognition processes and reduced user convenience.

Innovation Solution

The display device employs a speaker recognition system that identifies multiple speakers within a voice signal, separates their voice data, and processes these data differently based on the device's playback mode, allowing simultaneous or sequential execution of commands.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If conventional voice recognition technology is used to recognize user input, then the display device can respond to user commands, but it fails to accurately distinguish multiple speakers and is easily interfered with by nearby sound sources, leading to repeated recognition processes

Engineering Contradiction:
Improvevoice recognition accuracyVSAvoidspeaker identification reliability
Core Design Contradiction:
Measurement precisionVSReliability

Solution Approach 1:

The patent segments the voice recognition process by first identifying individual speakers through speaker recognition technology, then separating their voice data, and finally performing voice recognition on each speaker's data independently. This segmentation resolves the contradiction by enabling accurate distinction between multiple speakers while maintaining reliable recognition even in the presence of interference sounds.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent introduces speaker recognition as an intermediary step between voice input and voice recognition processing. This intermediary component identifies and separates multiple speakers before the actual voice recognition occurs, thereby improving both measurement precision of speaker identity and reliability of the overall recognition system in multi-speaker environments.

Inventive Principle:
Principle #24Intermediary (Mediator)

2Productivity

If the display device processes voice input from multiple speakers using conventional methods, then it can handle basic commands, but it performs repeated voice recognition processes when interference sounds are present, reducing operational efficiency

Engineering Contradiction:
Improvevoice processing efficiencyVSAvoidtime for repeated recognition processes
Core Design Contradiction:
ProductivityVSLoss of time

Solution Approach 1:

The patent applies preliminary action by performing speaker recognition and voice data separation before the main voice recognition process. By pre-identifying and isolating each speaker's voice data, the system avoids repeated recognition attempts caused by interference sounds, thereby improving processing efficiency and reducing time loss.

Inventive Principle:
Principle #10Preliminary action

3Adaptability or versatility

If the display device uses basic voice recognition without speaker identification, then the system remains simple, but it cannot provide personalized services or accurately match user intentions in multi-user scenarios

Engineering Contradiction:
Improvemulti-user service capabilityVSAvoidspeaker recognition system complexity
Core Design Contradiction:
Adaptability or versatilityVSDevice complexity

Solution Approach 1:

The patent implements multi-functionality by integrating speaker recognition, voice data separation, and voice recognition capabilities into a unified system. This allows the display device to handle both single-user and multi-user scenarios, providing personalized services while maintaining a cohesive system architecture that manages complexity effectively.

Inventive Principle:
Principle #6Universality (Multi-functionality)

Data Source

PatentEP4322538B1Display device and operating method thereof
Publication Date: 2025.07.30 LG ELECTRONICS INC
  • EP4322538B1 patent drawingFigure 1
  • EP4322538B1 patent drawingFigure 2
  • EP4322538B1 patent drawingFigure 3

AI summary

Disclosed are a display device and an operating method therefor. According to an aspect of the present disclosure, a method for operating a display device includes receiving voice data; separating the received voice data into pieces of voice data for speakers; and performing control such that pieces of content according to the pieces of voice data, which have been separated for the speakers are respectively output on multi-view screen areas, when a current playback mode is a multi-view mode.