Speech Reminder System Using Keyword Recognition for Time Extraction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current speech recognition technologies for setting reminders via speech input face challenges in accuracy, particularly in noisy environments and incomplete template matching, leading to incorrect or failed reminder settings due to instability in whole-text recognition and incomplete text sequence templates.
Innovation Solution
The method employs keyword recognition to accurately extract time information from speech signals, combined with continuous speech recognition to generate a text sequence, ensuring correct reminder settings even with incorrect text sequences, by using a dual-path recognition approach with foreground and background models to differentiate time-specific keywords and improve retrieval and recall rates.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If continuous speech recognition is used to recognize all speech signals to text sequence, then the system can process complete sentences, but the recognition accuracy becomes unstable especially in noisy environments
Solution Approach 1:
The patent divides the speech recognition task into two segments: keyword recognition for time information extraction and continuous speech recognition for text sequence generation. This segmentation allows the system to use the more accurate keyword recognition method specifically for time extraction, thereby resolving the accuracy instability problem while maintaining the ability to process complete sentences.
Solution Approach 2:
The patent applies partial action by using keyword recognition specifically for the critical time information extraction, rather than relying solely on continuous speech recognition for the entire sentence. This partial application of a more accurate method to the most important extraction task improves overall reliability.
2Productivity
If template matching is used to extract time information and reminder content, then the system can process structured speech, but incomplete template coverage causes extraction failures
Solution Approach 1:
The patent adds another dimension to the extraction process by combining template matching with keyword recognition. Instead of relying solely on template matching in the traditional dimension, the system incorporates keyword recognition as an additional dimension, allowing time information to be extracted through multiple pathways and thereby improving success rate.
Solution Approach 2:
The system makes the time extraction process universal by applying keyword recognition that can handle various time expression formats beyond what templates cover. This multi-functional approach allows the system to extract time information whether it matches predefined templates or uses alternative expressions.
3Measurement precision
If keyword recognition is used to extract time information, then extraction accuracy improves, but the system needs to differentiate time-specific keywords from other content
Solution Approach 1:
The patent segments the recognition system into distinct components: a keyword recognition module specifically for time information and a continuous speech recognition module for text sequence. This segmentation allows each module to be optimized for its specific task, improving accuracy while managing complexity through functional separation.
Solution Approach 2:
The system introduces an intermediary processing step where keyword recognition acts as a mediator between the speech signal and the final time extraction. This intermediary layer specifically targets time-related keywords, improving extraction accuracy while keeping the overall system architecture manageable through clear functional boundaries.
Data Source
AI summary
The present invention, pertaining to the field of speech recognition, discloses a reminder setting method and apparatus. The method includes: acquiring speech signals; acquiring time information in speech signals by using keyword recognition, and determining reminder time for reminder setting according to the time information; acquiring text sequence corresponding to the speech signals by using continuous speech recognition, and determining reminder content for reminder setting according to the time information and the text sequence; and setting a reminder according to the reminder time and the reminder content. According to the present invention, acquiring time information in speech signals by using keyword recognition ensures correctness of time information extraction, and achieves an effect that correct time information is still acquired by keyword recognition to set a reminder even in the case that a recognized text sequence is incorrect due to poor precision in whole text recognition in the speech recognition.