Recognition Apparatus Score Vector Filtering for Efficiency
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional recognition apparatuses face a decline in recognition rate when too many non-target symbols are skipped, leading to increased calculation costs due to reduced processing efficiency.
Innovation Solution
A recognition apparatus that filters score vector sequences by passing only score vectors with recognition-target symbols having the best scores, and those non-target symbols with scores worse than a threshold, while deleting vectors that do not meet specific conditions, thereby maintaining recognition accuracy with reduced calculation costs.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If non-target symbols are skipped to reduce calculation costs, then calculation efficiency is improved, but recognition rate deteriorates
Solution Approach 1:
The patent applies the skipping principle by selectively skipping non-target symbols in the score vector sequence. The filtering unit identifies and removes non-target symbols (symbols not belonging to the recognition target class) from the sequence, allowing the system to bypass unnecessary calculation steps while maintaining recognition accuracy for actual target symbols.
Solution Approach 2:
The patent extracts and removes non-target symbols from the score vector sequence through the filtering unit. This extraction process separates useful information (target symbols) from useless information (non-target symbols), reducing the computational burden on subsequent processing units while preserving the integrity of recognition-critical data.
2Reliability
If all score vectors are processed to maintain recognition accuracy, then recognition rate is improved, but calculation costs increase
Solution Approach 1:
The filtering unit extracts and removes non-target symbols from the score vector sequence, reducing the number of elements that subsequent processing units must handle. This extraction eliminates wasted computational resources on symbols that cannot contribute to recognition accuracy, thereby reducing overall calculation costs while maintaining effective processing of target symbols.
Solution Approach 2:
The patent applies partial action by processing only the necessary portion of the score vector sequence - specifically, only target symbols that can contribute to recognition. Instead of processing all symbols equally, the system performs partial processing on filtered sequences, reducing computational energy consumption while maintaining sufficient recognition accuracy.
Data Source
AI summary
According to an embodiment, a recognition apparatus includes one or more processors. The one or more processors are configured to calculate, based on the input signal, a score vector sequence in which a plurality of score vectors each including respective scores of symbols are arranged; and cause, among: a first score vector in which a representative symbol corresponding to a best score is a recognition-target symbol; a second score vector in which a representative symbol is a non-target symbol, and a score of the representative symbol is worse than a first threshold; and a third score vector in which a representative symbol is a non-target symbol, and a score of the representative symbol is equal to the first threshold or better than the first threshold, a third score vector satisfying a predefined first condition, to pass through to filter the score vector sequence.