Selective Audio Transmission for Speech Recognition Server Load
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The existing speech recognition systems face processing overload on external servers due to redundant data transmission, leading to decreased processing speed and increased unnecessary data transfer.
Innovation Solution
An electronic device equipped with a processor, microphone, speaker, and memory that compares local and external audio information to determine whether to transmit speech data to an external server, reducing redundant data transmission by selectively sending only necessary information.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If all electronic devices transmit speech information to the external server, then speech recognition service coverage is improved, but server processing load increases and processing speed decreases
Solution Approach 1:
The patent applies partial action by having only one electronic device (the one with the best audio quality) transmit speech information to the external server, rather than all devices transmitting. This selective transmission reduces server processing load while maintaining service coverage, as the server receives sufficient data from the optimal device to perform speech recognition.
Solution Approach 2:
The patent extracts the transmission function from all electronic devices and assigns it selectively to only the device with the best audio quality. By comparing audio quality metrics among devices and extracting the transmission task to the single best device, the system reduces redundant data transmission to the server while maintaining comprehensive service coverage.
2Reliability
If multiple electronic devices transmit speech information to the external server, then service reliability is improved, but data transmission redundancy increases
Solution Approach 1:
The system performs partial transmission by selecting only the necessary device (the one with best audio quality) to transmit speech information, rather than requiring all devices to transmit. This reduces unnecessary data transfer to the server while maintaining service reliability through selective optimal transmission.
Solution Approach 2:
The patent changes the transmission parameter from 'all devices transmit' to 'only the device with best audio quality transmits'. By evaluating audio quality parameters (such as signal-to-noise ratio, microphone quality, device proximity) and changing the transmission behavior based on these parameters, the system minimizes redundant data transfer while ensuring reliable speech recognition service.
Data Source
AI summary
Electronic devices and methods for controlling the electronic devices are provided. The electronic device includes a user interface, a communication circuit configured to connect to a cellular network and a Wi-Fi network, a microphone, a speaker, a processor connected to the user interface, the communication circuit, the microphone, and the speaker, and a memory connected to the processor and configured to store instructions executable by the processor for transmitting a first audio sound including first information associated with a wakeup utterance for calling a speech-based intelligence service through the speaker upon receipt of the wakeup utterance through the microphone, receiving a second audio sound including second information through the microphone, and determining whether to control the electronic device to transmit speech information following the wakeup utterance to an external server based on at least part of a comparison between the first information and the second information.