Speech Recognition Tree Structure for List Element Selection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current speech recognition systems are inefficient in selecting combinations of list elements from large databases using a single speech input, as they can only process a single element at a time, leading to computational inefficiencies, especially in applications like navigation systems that require selecting multiple elements simultaneously.
Innovation Solution
A speech recognition method that compares a speech input with two disjoint sets of list elements, generating candidate lists for each set and combining them to select the best matching combinations, utilizing techniques like wildcard symbols and scoring to enhance efficiency and reduce memory requirements.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If all valid combinations are enrolled into a single set of list elements, then the system can select combinations via speech input, but the processing efficiency deteriorates due to the large number of entries
Solution Approach 1:
The patent divides the single large set of all combinations into multiple smaller sets organized in a tree structure, where each node represents a set of list elements. This segmentation allows the system to process smaller subsets rather than the entire combination space, significantly improving processing efficiency while maintaining the ability to select combinations.
Solution Approach 2:
The patent introduces a hierarchical dimension by organizing sets in a tree structure with parent-child relationships. Instead of flattening all combinations into a single level, the system creates multiple levels of nesting, allowing efficient traversal and selection through the hierarchical structure rather than searching a flat large dataset.
2Adaptability or versatility
If a single set containing all combinations is used, then combination selection is enabled, but the memory requirements increase
Solution Approach 1:
The patent segments the large combination dataset into multiple smaller sets distributed across the tree structure. Each node contains only the list elements relevant to that particular branch, reducing the memory footprint at each processing step while maintaining access to all combinations through the hierarchical structure.
Solution Approach 2:
The patent extracts only the necessary subsets of list elements at each node based on the speech recognition progress. Instead of loading all combinations into memory simultaneously, the system extracts and processes relevant subsets dynamically, reducing overall memory requirements.
3Device complexity
If current speech recognition systems process single elements only, then processing is simpler, but the capability to select multiple elements simultaneously is lost
Solution Approach 1:
The patent segments the multi-element selection process into multiple single-element recognition steps, where each node in the tree represents a set of list elements to be recognized sequentially. This allows the system to leverage existing single-element speech recognition capabilities while achieving multi-element selection through hierarchical processing.
Solution Approach 2:
The patent creates a universal framework that handles both single-element and multi-element selection through the same tree structure. The system can select one element per node, and by traversing multiple nodes, it achieves multi-element selection capability, making the system versatile for different selection scenarios.
Data Source
AI summary
The invention provides a speech recognition method for selecting a combination of list elements via a speech input, wherein a first list element of the combination is part of a first set of list elements and a second list element of the combination is part of a second set of list elements, the method comprising the steps of receiving the speech input, comparing each list element of the first set with the speech input to obtain a first candidate list of best matching list elements, processing the second set using the first candidate list to obtain a subset of the second set, comparing each list element of the subset of the second set with the speech input to obtain a second candidate list of best matching list elements, and selecting a combination of list elements using the first and the second candidate list.