Speech Intent Prediction for Interaction Delay Reduction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing speech interaction technologies experience high delays due to the need for servers to interact with third-party resource servers to acquire response information, leading to prolonged processing times for user queries.
Innovation Solution
Implementing a method where a terminal device predicts user intentions in real-time or at preset intervals and sends these intentions to a server, allowing the server to acquire and store response information in advance, reducing the need for subsequent interactions with third-party servers.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If the terminal device waits to send complete speech information to the server after receiving it fully, then the accuracy of intention recognition is improved, but the speech interaction delay increases
Solution Approach 1:
The terminal device performs preliminary intention recognition on partial speech information during the reception process and sends predicted intentions to the server in advance. The server prepares response information based on these predictions before receiving the complete speech information, thereby reducing the overall interaction delay while maintaining accuracy through subsequent verification with the complete information.
2Loss of information
If the server interacts with third-party resource servers to acquire response information, then the comprehensiveness of response information is improved, but the processing time increases
Solution Approach 1:
The server performs preliminary actions by interacting with third-party resource servers in advance based on predicted speech intentions received from the terminal device. Response information is acquired and cached before the terminal device sends the complete speech information, significantly reducing the processing time when the complete information is received while ensuring response information completeness.
Solution Approach 2:
The server prepares and caches response information in advance as a cushioning measure. When the complete speech information is received, the server can directly use the pre-prepared response information or make minimal adjustments, avoiding the need to interact with third-party servers at the moment of complete information reception, thus reducing processing time.
3Loss of time
If the terminal device sends speech intentions during the reception process, then the response time is reduced, but the complexity of intention recognition increases
Solution Approach 1:
The speech information reception process is segmented into multiple stages. The terminal device performs intention recognition on partial speech information at different segments during the reception process, sending predictions to the server progressively. This segmentation allows the system to start processing early while managing the complexity by handling smaller portions of information at each stage.
Data Source
AI summary
The present disclosure provides a speech processing method. A specific implementation solution is: a terminal device sends at least one speech intention to a server in a process of receiving first speech information, where each speech intention is a speech intention corresponding to a part of speech information in the first speech information; the server acquires response information corresponding to the at least one speech intention; the terminal device sends the first speech information to the server in response to completion of receiving the first speech information; the server acquires a second speech intention corresponding to the first speech information, and sends the response information corresponding to the first speech intention to the terminal device, and the terminal device outputs the response information.