Speech Recognition Audio Phrase Selection Buffer
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current technologies lack effective methods for selecting key phrases from audio contents, which are essential for efficient interaction with audio information, especially in mobile devices where visual representation is limited.
Innovation Solution
A system and method utilizing speech recognition to identify specific audio phrases from a series of audio contents by performing speech recognition between the audio contents and user input, generating a corresponding textual phrase for further processing, including the use of a buffer to store audio and speech input within a desired period and employing techniques like hidden Markov models for recognition.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If speech recognition is used to select audio contents, then user interaction efficiency is improved, but system complexity increases
Solution Approach 1:
The patent introduces a buffer as an intermediary component that temporarily stores audio contents and speech inputs. This buffer mediates between the audio playback system and the speech recognition system, enabling the recognition system to process audio segments without requiring direct access to the entire audio stream, thus reducing overall system complexity while maintaining interaction efficiency
Solution Approach 2:
The patent segments the audio contents into manageable buffers that can be processed independently by the speech recognition system. Instead of processing the entire audio stream at once, the system divides audio into temporal segments stored in the buffer, allowing more efficient and less complex processing of speech patterns within specific time windows
2Loss of information
If audio contents are processed sequentially, then information representation is improved, but key phrase selection becomes difficult
Solution Approach 1:
The patent implements feedback by comparing speech recognition results with the original audio contents. The system uses the buffer to store both the audio segments and the corresponding speech inputs, then feeds this information back to identify and highlight key phrases. This feedback mechanism enables the system to distinguish important phrases from generic audio content while maintaining the sequential nature of audio processing
Data Source
AI summary
A system and method for selecting audio contents by using the speech recognition to obtain a textual phrase from a series of audio contents are provided. The system includes an output module outputting the audio contents, an input module receiving a speech input from a user, a buffer temporarily storing the audio contents within a desired period and the speech input, and a recognizing module performing a speech recognition between the audio contents within the desired period and the speech input to generate an audio phrase and the corresponding textual phrase matching with the speech input.