Speech Input Mode Switching for Dictation and Voice Assistant Tasks

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Electronic devices often provide inconsistent results for the same user's utterance depending on the speech processing method, struggling to accurately process user utterances in different modes such as dictation or conversation services.

Innovation Solution

The electronic device includes a processor and memory configured to execute operations based on user input methods, selectively engaging an external server for automatic speech recognition (ASR) and intelligence systems, allowing for various user utterance processing modes based on device state or input type.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Adaptability or versatility

If the electronic device uses a single speech processing method, then the processing flow is simple, but the device cannot adapt to different user needs (dictation vs. conversation services)

Engineering Contradiction:
Improveadaptability to different speech processing modesVSAvoidcomplexity of speech processing system
Core Design Contradiction:
Adaptability or versatilityVSDevice complexity

Solution Approach 1:

The system dynamically switches between different speech processing modes (dictation mode and conversation mode) based on user input characteristics. The processor determines whether to perform natural language understanding based on the presence of question words, enabling adaptive processing without requiring multiple fixed systems.

Inventive Principle:
Principle #15Dynamics

Solution Approach 2:

A single speech processing system performs multiple functions by selectively applying different processing methods. The same processor handles both direct text input and speech recognition, and conditionally applies natural language understanding only when needed, making the system universal rather than requiring separate dedicated systems.

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

2Reliability

If the device always performs natural language understanding on speech input, then conversation services are improved, but processing time and computational resources increase unnecessarily for simple dictation tasks

Engineering Contradiction:
Improveaccuracy of speech processingVSAvoidspeech processing time
Core Design Contradiction:
ReliabilityVSLoss of time

Solution Approach 1:

The system changes the processing parameter (whether to perform natural language understanding) based on the input characteristics. When question words are detected in speech input, the system switches to conversation mode with full NLU processing; otherwise, it uses simpler dictation mode processing, optimizing both accuracy and speed for different scenarios.

Inventive Principle:
Principle #35Parameter changes

Solution Approach 2:

Natural language understanding is applied selectively only to the portions of speech input that require it (those containing question words), rather than uniformly to all input. This localized application of complex processing improves accuracy where needed while maintaining efficiency elsewhere.

Inventive Principle:
Principle #3Local quality

3Measurement precision

If the device processes all speech inputs with full natural language understanding, then processing accuracy is improved, but the device complexity and computational load increase

Engineering Contradiction:
Improveprecision of speech interpretationVSAvoidcomplexity of processing algorithms
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The system performs partial natural language understanding only when necessary (detecting question words), rather than applying full NLU processing to all inputs. This partial action approach maintains precision for conversation queries while avoiding the computational overhead of universal complex processing.

Inventive Principle:
Principle #16Partial or excessive action

4Productivity

If the device switches processing modes based on device state, then processing efficiency is improved, but the control logic becomes more complex

Engineering Contradiction:
Improvespeech processing efficiencyVSAvoidcomplexity of mode switching control
Core Design Contradiction:
ProductivityVSDevice complexity

Solution Approach 1:

The system uses simple parameter changes (detecting the presence of question words in the input) to trigger mode switches between dictation and conversation processing. This straightforward parameter-based control achieves efficient mode switching without requiring complex state machine logic or multiple control layers.

Inventive Principle:
Principle #35Parameter changes

Data Source

PatentEP4137934B1Electronic device for processing user speech and operating method therefor
Publication Date: 2025.12.03 SAMSUNG ELECTRONICS CO LTD
  • EP4137934B1 patent drawingFigure 1
  • EP4137934B1 patent drawingFigure 2
  • EP4137934B1 patent drawingFigure 3

AI summary

An electronic device of the present invention comprises: a housing; a touchscreen display; a microphone; at least one speaker; a button disposed on a portion of the housing or set to be displayed on the touchscreen display; a wireless communication circuit; a processor; and a memory. The electronic device is configured to store an application program including a user interface for receiving a text input. When the user interface is not displayed on the touchscreen display, the electronic device enables a user to receive a user input through the button, receives user speech through the microphone, and then provides data on the user speech to an external server including an automatic speech recognition system and an intelligence system. An instruction for performing a task generated by the intelligence system in response to the user speech is received from the server. When the user interface is displayed on the touchscreen display, the electronic device enables the user to receive the user input through the button, receives user speech through the microphone, and then provides data on the user speech to the external server. Data on the text generated by the automatic speech recognition is received from the server, but the instruction generated by the intelligence system is not received therefrom. Other embodiments are also possible.