Speaking User Selection Using Random Models in Group Chat

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing multi-user chat room systems lack flexibility in selecting speaking users, leading to fixed and less engaging social activities due to manual manager intervention, which affects the smoothness and fun of interactions.

Innovation Solution

A method and apparatus for randomly selecting a speaking user using a random selecting model, allowing automatic determination of a speaking user based on a user group, enhancing flexibility and engagement in social activities.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If multiple microphones are used to pick up speech in different directions, then speech recognition accuracy is improved, but determining which microphone corresponds to the speaking user becomes more difficult

Engineering Contradiction:
Improvespeech recognition accuracyVSAvoidmicrophone selection complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The system uses feedback from the audio processing unit to analyze speech characteristics and directional information from each microphone, then feeds this information back to the controller to determine which microphone corresponds to the speaking user. This feedback loop resolves the contradiction by providing intelligent selection based on real-time analysis rather than simple sequential checking.

Inventive Principle:
Principle #23Feedback

Solution Approach 2:

The patent replaces the mechanical/sequential approach of determining microphone correspondence with an information-processing approach. Instead of physically or sequentially testing each microphone, the system uses the audio processing unit to analyze speech characteristics, directionality, and other parameters to computationally determine which microphone received the speech, thereby simplifying the selection process while maintaining accuracy.

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

2Measurement precision

If sequential speech recognition is performed for each microphone, then speaking user selection accuracy is improved, but recognition time is excessively long

Engineering Contradiction:
Improvespeaking user selection accuracyVSAvoidrecognition time
Core Design Contradiction:
Measurement precisionVSLoss of time

Solution Approach 1:

The system performs preliminary analysis by the audio processing unit on speech characteristics, directionality, and other parameters from all microphones simultaneously before the final recognition decision. This preliminary action prepares the data in advance, allowing the controller to quickly determine the speaking user without performing full sequential recognition on each microphone, thereby reducing recognition time while maintaining accuracy.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The patent segments the speech recognition process into two parts: (1) preliminary analysis of speech characteristics and directional information by the audio processing unit for all microphones, and (2) final speaking user determination by the controller based on this pre-processed information. This segmentation allows parallel processing of initial data collection while maintaining sequential decision-making for accurate user identification, thus reducing overall recognition time.

Inventive Principle:
Principle #1Segmentation

3Measurement precision

If the system waits for all users to finish speaking before recognition, then accuracy is improved, but real-time responsiveness is lost

Engineering Contradiction:
Improvespeech recognition accuracyVSAvoidreal-time responsiveness
Core Design Contradiction:
Measurement precisionVSSpeed

Solution Approach 1:

The system dynamically adjusts its recognition approach based on real-time conditions. The controller can initiate speaking user determination as soon as speech is detected and analyzed by the audio processing unit, rather than waiting for a fixed period or all users to finish. This dynamic timing allows the system to balance accuracy with real-time responsiveness, recognizing speech promptly when conditions are favorable while maintaining the ability to wait if needed for better accuracy.

Inventive Principle:
Principle #15Dynamics

Data Source

PatentEP4346171B1Speaking user selecting method and apparatus, electronic device, and storage medium
Publication Date: 2026.05.20 DOUYIN VISION CO LTD
  • EP4346171B1 patent drawingFigure 1
  • EP4346171B1 patent drawingFigure 2~3
  • EP4346171B1 patent drawingFigure 4

AI summary

A speaking user selecting method and apparatus, an electronic device, a storage medium, a computer program product, and a computer program. The method comprises: receiving a first request sent by a first terminal device, the first request being used for determining a speaking user in a user group, wherein the user group comprises at least two users to be selected, and each user to be selected corresponds to one second terminal device; determining a random selection model corresponding to the first request, the random selection model being used for representing a policy for randomly determining a speaking user from the user group; and determining a speaking user from the at least two users to be selected according to the random selection model, and sending identification information of the speaking user to the first terminal device and the second terminal device. In this process, an administrator does not need to manually select a speaking user, so that the fluency and interestingness of a social activity implementation process based on a multi-person chat group can be improved in a specific social activity scene, and the flexibility and diversity of social activity implementation modes are expanded.