Multi-View Display Voice Separation for Speaker-Specific Content
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional voice recognition technologies in display devices struggle with accurately distinguishing multiple speakers and handling interference sounds, leading to inefficient and inconvenient user interactions.
Innovation Solution
The display device employs a method to identify multiple speakers by separating voice data for each speaker and performing Speech-to-Text (STT) and Natural Language Processing (NLP) on each voice data segment, allowing content output to corresponding screen areas based on the current playback mode, particularly in a multi-view mode.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional voice recognition technology is used, then the display device can respond to user commands, but it cannot accurately distinguish multiple speakers and is affected by interference sounds
Solution Approach 1:
The patent segments the mixed audio signal into separate speaker channels using speaker separation technology. The audio processing unit divides the composite voice data from multiple speakers into individual speaker voice data, enabling the system to identify and process each speaker's commands independently, thereby resolving the contradiction between recognition accuracy and multi-speaker adaptability
Solution Approach 2:
The patent introduces an audio processing unit as an intermediary component between the microphone array and the voice recognition system. This intermediary performs speaker separation and identification, acting as a mediator that transforms the complex multi-speaker audio signal into structured, speaker-specific data that the recognition system can accurately process
2Reliability
If voice recognition is performed repeatedly due to interference sound recognition, then the system attempts to handle uncertain inputs, but user convenience deteriorates due to repeated processing
Solution Approach 1:
The patent performs preliminary speaker separation and identification before voice recognition processing. By pre-processing the audio signal to isolate individual speakers and validate voice data quality, the system avoids repeated recognition attempts and reduces unnecessary processing time, thereby improving reliability without sacrificing efficiency
3Ease of operation
If multi-view mode is implemented with speaker-specific content output, then user convenience increases through targeted content delivery, but device complexity increases due to additional processing requirements
Solution Approach 1:
The patent implements a universal audio processing framework that handles both single-speaker and multi-speaker scenarios through the same speaker separation and identification mechanisms. The system universally applies speaker-specific content matching to control operations, whether in single-view or multi-view mode, thereby improving user convenience without proportionally increasing complexity
Data Source
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 a plurality of speakers; and performing control such that pieces of content respectively corresponding to the pieces of voice data, which have been separated for the plurality of speakers, are respectively output on corresponding screen areas of multi-view screen areas, when a current playback mode is a multi-view mode.