Erroneous Conversion Dictionary Creation for Voice Recognition
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The error-tendency-learning voice recognition device requires complex calculations to grasp error tendencies, leading to a complicated process.
Innovation Solution
An incorrect conversion dictionary generating system that converts input terms to voice data, performs voice analysis, and determines incorrectly converted terms, which are then stored in a dictionary for use in voice recognition systems.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If error correction model calculations are performed to grasp error tendency, then voice recognition accuracy is improved, but process complexity increases
Solution Approach 1:
The patent applies preliminary action by pre-collecting and storing erroneously converted terms in a dictionary before actual voice recognition occurs. During voice recognition, the system simply checks against this pre-prepared dictionary rather than performing complex error tendency calculations in real-time, thus improving speed while maintaining accuracy.
Solution Approach 2:
The patent creates a simplified copy of error patterns by collecting actual erroneous conversions and storing them in a dictionary. Instead of using complex error correction models, the system copies real-world error data and uses it for quick comparison during voice recognition, reducing computational complexity while preserving recognition accuracy.
2Measurement precision
If complex error correction model calculations are performed, then error tendency is grasped accurately, but processing time increases
Solution Approach 1:
The system performs error data collection and dictionary creation in advance before actual voice recognition tasks. This preliminary action allows the system to have error tendency information ready without performing complex calculations during time-critical voice recognition processing, thus maintaining accuracy while reducing processing time.
Solution Approach 2:
The patent copies actual error conversion data into a searchable dictionary format, creating a simplified representation of error patterns. This copied error data can be quickly matched during voice recognition without re-performing complex error tendency calculations, thereby maintaining measurement precision while significantly reducing processing time.
Data Source
AI summary
An incorrect conversion dictionary generating system includes a term input unit, a voice conversion unit 5 that converts an input term to voice to obtain input-term voice, a voice analysis unit that receives the input-term voice output from the voice conversion unit, performs a voice analysis to convert the input-term voice to a term, and obtains a voice analyzed term, and an incorrectly converted term determining unit that receives the input term from the term input unit or the voice conversion unit, receives the voice analyzed term from the voice analysis unit, and determines the voice analyzed term as an incorrectly converted term of the input term when the input term does not match the voice analyzed term. The incorrect conversion dictionary generating system receives the input term and the incorrectly converted term, associates the input term with the incorrectly converted term, and stores in an incorrect conversion dictionary.