Speech Recognition Proper Noun Correction via Personalized Database

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Speech recognition systems in electronic devices often fail to accurately interpret terms with high error potential, such as proper nouns or user-created coined words, leading to incorrect command execution and mismatched user intent.

Innovation Solution

An electronic device equipped with a microphone, display, and processor that executes a speech recognition function, displays generated text, and provides recommendations from a personalized database for recognized proper nouns, allowing users to revise errors and ensure accurate command execution.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If a standard speech recognition system is used, then the device can process common commands, but it fails to accurately recognize proper nouns and user-created coined words

Engineering Contradiction:
Improvespeech recognition accuracyVSAvoidhandling of proper nouns and coined words
Core Design Contradiction:
Measurement precisionVSAdaptability or versatility

Solution Approach 1:

The system pre-processes user utterances by identifying proper nouns and coined words before main recognition, preparing candidate corrections in advance from the personalized database. This preliminary action allows the system to address difficult recognition cases proactively rather than reactively.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The system provides feedback to users by displaying candidate correction terms when recognition uncertainty is detected. Users can select from suggested corrections or provide manual input, and this feedback loop enables continuous improvement of recognition accuracy for proper nouns and coined words.

Inventive Principle:
Principle #23Feedback

2Measurement precision

If the system displays candidate corrections for proper nouns, then recognition accuracy improves, but the interaction time increases

Engineering Contradiction:
Improveterm recognition accuracyVSAvoiduser interaction time
Core Design Contradiction:
Measurement precisionVSLoss of time

Solution Approach 1:

The system applies partial action by only displaying candidate corrections when recognition uncertainty exceeds a threshold, rather than always showing corrections. This selective approach reduces unnecessary interaction steps while maintaining accuracy for problematic terms.

Inventive Principle:
Principle #16Partial or excessive action

Solution Approach 2:

The system allows users to skip the correction selection process by providing a default option or automatic selection capability, enabling quick passage through the correction step when the user is confident in the original recognition or when time is critical.

Inventive Principle:
Principle #21Skipping (Rushing through)

3Measurement precision

If a personalized database is maintained for each user, then recognition of user-specific terms improves, but device complexity increases

Engineering Contradiction:
Improveuser-specific term recognitionVSAvoidpersonalized database management
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The personalized database system is designed to be universal across different users and device types. The same database structure and management mechanisms serve multiple users and applications, reducing overall system complexity through standardized multi-functional design.

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

Solution Approach 2:

The system automatically updates and maintains the personalized database without requiring manual user configuration. It self-learns from user corrections and interactions, automatically incorporating new proper nouns and coined words into the database, thereby reducing the burden of database management.

Inventive Principle:
Principle #25Self-service

Data Source

PatentEP3709294B1Electronic device for providing speech recognition service and method thereof
Publication Date: 2021.08.25 SAMSUNG ELECTRONICS CO LTD
  • EP3709294B1 patent drawingFigure 1
  • EP3709294B1 patent drawingFigure 2
  • EP3709294B1 patent drawingFigure 3

AI summary

An electronic device is provided. The electronic device includes a microphone, a display, a memory, and a processor electrically connected to the microphone, the display, and the memory. The processor is configured to execute a speech recognition function, to obtain a first utterance from a user through the microphone, to display a text generated based on the first utterance, in the display, and to display at least one item, which corresponds to a portion recognized as a proper noun in the text and which is determined based on the first utterance and a personalized database of the user, in the display.