Speech Recognition Using Historical User Data
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing speech recognition systems are often inaccurate due to factors like clarity of speech, background noise, and systemic limitations, leading to user frustration and reduced adoption in applications like portable devices and self-service kiosks.
Innovation Solution
The integration of historical information into speech recognition systems, where user-specific data is retrieved and used to modify recognition results, improving accuracy by leveraging past interactions and preferences.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If speech recognition is performed without historical information, then the system operates with simpler processing, but the recognition accuracy is reduced
Solution Approach 1:
The system performs preliminary actions by retrieving and storing historical speech recognition data and user preference information before the actual recognition task. This historical information is prepared in advance and made available to enhance the accuracy of current recognition without adding complex real-time processing requirements.
Solution Approach 2:
Historical information acts as an intermediary between the speech input and the recognition output. The system retrieves past recognition results and user preferences, uses them to disambiguate current speech inputs, and produces more accurate results. This intermediary layer of historical data mediates the recognition process without requiring fundamental changes to the core recognition engine.
2Measurement precision
If historical information is retrieved and used to modify recognition results, then recognition accuracy is improved, but system response time increases
Solution Approach 1:
Historical speech recognition data and user preference information are retrieved and stored in advance before the actual recognition task occurs. This preliminary preparation of historical information allows the system to quickly access relevant data during recognition without adding significant real-time processing delays.
Solution Approach 2:
The system retrieves only the necessary portion of historical information relevant to the current speech input rather than processing all historical data. By selectively accessing only the relevant historical recognition results and user preferences needed for the current task, the system minimizes data retrieval time while still achieving improved accuracy.
Data Source
AI summary
Embodiments of the present invention improve speech recognition using historical information. In one embodiment, the present invention includes a method of performing speech recognition comprising receiving an identifier specifying a user of a kiosk, retrieving history information about the user using the identifier, receiving speech input, recognizing said speech input in the context of a first recognition set, resulting in first recognition results, and modifying the first recognition results using the history information.