Voice Recognition Dictionary Clustering for Unknown Word Registration
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current voice recognition systems fail to effectively recognize and improve the recognition rate for repeatedly inputted unknown words, as they are not registered in the voice word dictionary until manually added, limiting their ability to recognize similar unknown words.
Innovation Solution
A voice processing circuit and method that utilizes dynamic programming to extract phonemes and syllables, clusters unknown words based on characteristic distances, and updates the dictionary by registering frequently occurring unknown words as known words, improving recognition rates for similar inputs.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If unknown words are not registered in the voice word dictionary, then the system maintains a clean and manageable dictionary, but the recognition rate for unknown words remains at zero
Solution Approach 1:
The voice recognition system automatically detects unknown words, clusters similar ones together based on phoneme patterns, and registers representative words into the dictionary without manual intervention. This self-service mechanism allows the system to autonomously improve its recognition rate for unknown words while maintaining dictionary organization through automated clustering algorithms.
Solution Approach 2:
The system performs preliminary clustering of unknown words based on phoneme sequences before registration. By pre-grouping similar unknown words and selecting representatives, the system prepares the dictionary in advance, ensuring that future similar unknown words can be recognized through the clustered patterns, thus improving recognition rate proactively.
2Reliability
If the system processes each unknown word individually, then the processing logic remains simple, but the recognition rate for repeatedly input similar unknown words does not improve
Solution Approach 1:
The system merges multiple similar unknown words into clusters based on phoneme sequence similarity. By combining words with similar phonetic patterns into groups and selecting representative words for registration, the system improves recognition rate for all words in the cluster while managing complexity through systematic clustering rather than individual processing.
Solution Approach 2:
The system changes the parameter of word representation from individual word forms to phoneme-based clusters. By transforming unknown words into phoneme sequences and grouping them by similarity metrics, the system creates a more efficient representation that improves recognition rate without requiring complex individual processing logic for each word variant.
3Reliability
If the dictionary is updated with every unknown word encountered, then the recognition rate improves continuously, but the dictionary size and maintenance burden increase significantly
Solution Approach 1:
Instead of adding every unknown word individually to the dictionary, the system merges similar unknown words into clusters and registers only representative words. This merging approach maintains recognition rate improvement while significantly reducing dictionary size by eliminating redundant entries of phonetically similar words.
Solution Approach 2:
The system discards redundant unknown word entries by clustering them with similar words, keeping only the representative form in the dictionary. When similar unknown words are encountered again, the system recovers recognition accuracy through the clustered pattern matching, thus maintaining recognition rate without increasing dictionary size with duplicate entries.
Data Source
Figure 1
Figure 2
Figure 3A~3C
AI summary
A voice processing apparatus includes a first storage unit (14A) which stores a known-word, a voice recognition unit (13) which extracts an unknown-word by executing a voice recognition process on an input voice signal, based on a storage content of the first storage unit (14A), and a storage control unit (13') which executes storage control to the first storage unit (14A), wherein the storage control unit (13') stores, when information of a number of unknown-words which are recognized to be identical, among unknown-words extracted by the voice recognition unit (13), meets a predetermined condition, a corresponding unknown-word in the first storage unit (14A) as a known-word.