Gateway Voice Feature Extraction for Smart Home Privacy
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional voice recognition systems for controlling household electrical appliances transmit all voice information to a server, leading to data concentration and privacy concerns, as well as increased network load, without adequate consideration for user privacy or efficient data transmission.
Innovation Solution
A control method involving a gateway that analyzes voice information and transmits only relevant data to a server for recognition, allowing the gateway to execute control instructions independently for common phrases and reducing data transmission by transmitting only necessary information to the server, while also obtaining user confirmation for control actions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If all voice information is transmitted to the server for recognition, then voice recognition accuracy is improved, but data transmission volume increases and user privacy is compromised
Solution Approach 1:
The patent extracts and transmits only the essential feature parameters of voice information (such as MFCC coefficients) to the server, rather than transmitting the complete raw voice data. This extraction approach maintains sufficient information for accurate voice recognition while significantly reducing the data transmission volume and protecting user privacy.
Solution Approach 2:
The voice information processing is segmented into two stages: local feature extraction at the client device and remote recognition at the server. This segmentation allows the client to perform preliminary processing and send only necessary extracted features, reducing overall data transmission while maintaining recognition accuracy.
2Reliability
If all voice information is transmitted to the server, then recognition completeness is improved, but network load increases
Solution Approach 1:
The patent extracts essential feature parameters from voice information before transmission, removing redundant data while retaining the core information needed for recognition. This reduces network load significantly while maintaining recognition completeness.
Solution Approach 2:
The client device performs preliminary feature extraction and preprocessing of voice information before transmission to the server. This preliminary action reduces the amount of data that needs to be transmitted over the network while ensuring that the server receives sufficient information for complete recognition.
3Quantity of substance
If voice information is compressed-encoded before transmission, then data transmission volume is reduced, but processing capability requirements increase
Solution Approach 1:
The client device performs self-service by extracting and encoding voice features locally before transmission. This self-processing approach reduces the burden on the server and network while maintaining efficient data transmission, as the encoding is performed using standardized algorithms that balance compression efficiency with computational feasibility.
Data Source
AI summary
A gateway (103) determines whether or not voice information collected by a sound collecting device (102) needs to be subjected to voice recognition by a server apparatus (101), and only voice information that is determined to need to be subjected to voice recognition by the server apparatus (101) is transmitted by the gateway (103) to the server apparatus (101). The server apparatus (101) recognizes the received voice information, decides a control instruction, and transmits the control instruction to a household electrical appliance (104) via the gateway (103).