Distance-Weighted Speech Recognition Across Multiple Terminals
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current speech recognition technologies experience significant accuracy drops when the distance between the user and the target device increases, limiting the effectiveness of voice-controlled device interactions in indoor settings.
Innovation Solution
A speech recognition server calculates the time of arrival of speech sounds at multiple terminals, determines distances between the user and these terminals, and integrates speech recognition results using weighted values based on these distances to improve recognition accuracy across varying distances.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If speech recognition is performed using a single terminal device, then the system complexity is low, but the recognition accuracy decreases significantly as the distance between user and target device increases
Solution Approach 1:
The system divides the speech recognition function across multiple terminal devices distributed in different locations. Each terminal captures speech signals independently, and the server integrates results from multiple sources to maintain high recognition accuracy regardless of distance between user and target device.
Solution Approach 2:
The server combines speech recognition results from multiple terminals using weighted integration based on distance metrics. By merging data from multiple sources, the system achieves robust recognition accuracy while managing complexity through centralized processing.
2Reliability
If multiple terminals are used to improve speech recognition accuracy at distance, then the recognition accuracy improves, but the device complexity and computational load increase
Solution Approach 1:
The system applies different weighting factors to speech recognition results from different terminals based on their respective distances to the user. Terminals closer to the user contribute more to the final recognition result, optimizing reliability while managing computational complexity through selective emphasis on high-quality inputs.
Solution Approach 2:
The system dynamically adjusts weighting parameters for each terminal based on calculated distances. This parameter adaptation allows the system to maintain high reliability by emphasizing inputs from optimally positioned terminals while keeping the overall system complexity manageable through systematic parameter management.
3Ease of operation
If the user must be physically close to the target device for accurate speech recognition, then the recognition accuracy is high, but the ease of operation and user convenience decrease
Solution Approach 1:
The system enables speech recognition control of any target device from any location within the network environment. Multiple terminals serve as both speech capture devices and communication nodes, allowing users to control devices remotely without physical proximity constraints, thus achieving universal accessibility.
Solution Approach 2:
The server acts as an intermediary that receives speech signals from multiple terminals, integrates the recognition results using distance-based weighting, and transmits commands to target devices. This intermediary architecture enables accurate speech recognition and device control regardless of the physical distance between user and target device.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
This approach enables accurate voice command recognition and control of devices regardless of distance, allowing for seamless interaction with multiple devices through a networked system without the need for the user to be physically close to the target device.
Implementation Method 1
calculate times of arrival of a speech sound at each of the terminals using speech signals received from each of the terminals
Implementation Method 2
calculate distances between a user and the terminals based on the times of arrival of the speech sound to each of the terminals
Data Source
AI summary
A speech recognition server and method are provided. The speech recognition server includes: a network connector configured to connect to terminals through a communication network, and a processor configured to calculate times of arrival of a speech sound at each of the terminals using the speech signals received from each of the terminals, calculate distances between a user and the terminals based on the times of arrival of the speech sound to each of the terminals, and recognize a command to control a target by integrating speech recognition results obtained from the speech signals using a weighted value determined based on the distances.