Phoneme Buffering for Speech Recognition During Active Audio
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Electronic devices face difficulties in performing speech recognition when audio is being received or played during call or recording applications, as they typically require a designated input to activate the speech recognition service.
Innovation Solution
An electronic device acquires and stores phoneme information from audio data while a designated application is running, and transmits this information to an external device when a specific condition is met, allowing the external device to generate control information for executing related applications based on the stored phoneme information.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If the electronic device activates speech recognition application based on wake-up method (call word input, touch, or button), then the speech recognition service can be started, but it cannot perform speech recognition on speech uttered while audio is being received through call application or while audio is being played through recording application
Solution Approach 1:
The electronic device stores phoneme information from audio data in advance (before speech recognition is triggered) in a buffer or memory. When a designated input is received, the pre-stored phoneme information is immediately used for speech recognition, eliminating the need to wait for wake-up activation. This allows speech recognition to occur seamlessly during audio application execution without interrupting the audio flow.
2Use of energy by moving object
If the electronic device performs speech recognition only after activating speech recognition application, then the system resource consumption is controlled, but the speech recognition response time is delayed and cannot recognize speech before activation
Solution Approach 1:
Instead of performing full speech recognition processing continuously (which would consume excessive resources), the device performs partial processing by extracting and storing only phoneme information from the audio data in real-time. The complete speech recognition is then performed using this pre-prepared phoneme information when needed, achieving fast response with minimal ongoing resource consumption.
3Productivity
If the electronic device stores phoneme information from audio data, then speech recognition can be performed on audio received before designated input, but additional memory storage and data processing are required
Solution Approach 1:
The device extracts only the essential phoneme information from the audio data stream, rather than storing or processing the entire audio data. This extraction approach captures the critical speech elements needed for recognition while significantly reducing the storage burden and processing complexity compared to handling complete audio files.
Data Source
AI summary
An electronic device according to various embodiments may comprise a memory in which one or more applications are installed, a communication circuit, and a processor, wherein the processor is configured to acquire audio data during execution of a designated application among the one or more applications, wherein the acquiring of audio data comprises an operation of storing, in the memory, at least a portion including multiple pieces of phoneme information among the audio data, when a designated condition is satisfied, transmit the at least portion to an external electronic device so that the external electronic device generates designated information for execution of at least one application among the one or more applications by using at least a part of the multiple pieces of phoneme information stored before the designated condition is satisfied, and on the basis of the designated information, execute the at least one application in relation to the designated application.