Speech Matching Device Using Approximate Pinyin Mapping
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current speech input matching technologies have a low success rate due to pronunciation differences between standard and regional dialects and confusability of certain sounds in spoken languages, leading to mismatched text outputs.
Innovation Solution
A method and device that generate approximate pinyin for speech input based on pronunciation similarity information, using a preset mapping relationship table to improve matching accuracy by allowing fuzzy matching between standard and non-standard pronunciations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional speech recognition is used, then the system is simple, but the matching success rate is low due to dialectal pronunciation differences
Solution Approach 1:
The system pre-establishes a mapping relationship table that contains approximate pinyin for common pronunciation errors before speech recognition occurs. When speech input is received, the system queries this pre-prepared table to find matching words, avoiding the need for complex real-time pronunciation adjustment while improving matching accuracy for dialectal variations.
Solution Approach 2:
The patent introduces an intermediary mapping relationship table that bridges the gap between standard pinyin and dialectal pronunciations. This table acts as a mediator by storing correspondence between standard words and their approximate pinyin representations, allowing the system to translate dialectal speech patterns into standard text without requiring complex real-time processing.
2Adaptability or versatility
If strict pinyin matching is used, then the text output is precise, but it fails to accommodate non-standard pronunciations from different dialects
Solution Approach 1:
The system changes the matching parameter from exact pinyin equality to approximate pinyin matching based on the mapping relationship table. By allowing parameter variation in the pinyin representation (accepting approximate matches rather than exact matches), the system becomes adaptable to dialectal pronunciations while maintaining reasonable text output accuracy through the pre-established correspondence rules.
Data Source
AI summary
A method and device for matching speech to text are disclosed, the method including: receiving a speech input, the mentioned speech input carrying input speech information; obtaining initial text corresponding to the input speech information, and respective pinyin of the initial text; generating at least one approximate pinyin for the initial text based on predetermined pronunciation similarity information; and from a preset mapping relationship table, obtaining additional text corresponding to the respective pinyin of the initial text or to the at least one approximate pinyin of the initial text, wherein the preset mapping relationship table includes a respective record for each word in a word database, including respective pinyin and at least one respective approximate pinyin for said each word, and a respective mapping relation between said respective pinyin, said at least one respective approximate pinyin, and said each word.