Interactive Voice Synthesis Using Real-Time Speech Flow Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing voice synthesis methods lack true interactivity and adaptability, requiring human intervention or pre-established scripts, limiting flexibility and audience engagement.
Innovation Solution
A method for automatically selecting a group of words in a text based on real-time speech analysis to seamlessly integrate voice synthesis with human speech, allowing for intelligent audio alternation without pre-scripted interventions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If manual intervention is used to control audio alternation between human speech and voice synthesis, then the system can handle complex interactions, but the ease of operation deteriorates due to requiring human intervention
Solution Approach 1:
The system automatically detects when the speaker has finished speaking and triggers voice synthesis without requiring manual intervention. The speaker simply speaks naturally, and the system self-manages the transition to synthesized speech, eliminating the need for buttons, gestures, or voice commands to activate the synthesis.
Solution Approach 2:
The system uses real-time analysis of captured speech to detect when the speaker has finished, creating a feedback loop that automatically triggers the appropriate response. This feedback mechanism allows the system to adapt to the speaker's natural rhythm without pre-established scripts or manual control.
2Extent of automation
If pre-established parameterization is used for voice synthesis, then the extent of automation improves, but the adaptability deteriorates due to limited interactivity
Solution Approach 1:
The system dynamically adjusts the voice synthesis based on real-time detection of when the speaker has finished, rather than following a fixed pre-established script. This dynamic approach allows the system to adapt to varying speech patterns and pauses while maintaining full automation.
Solution Approach 2:
The system performs preliminary analysis of the captured speech to determine when the speaker has finished, enabling automatic triggering of voice synthesis before manual intervention would be required. This preliminary detection allows the system to be both automated and adaptive.
3Adaptability or versatility
If voice synthesis is triggered by speech interruption detection, then the adaptability improves, but the reliability deteriorates due to potential false detection
Solution Approach 1:
The system applies preliminary analysis to the captured speech to distinguish between intentional speech interruptions (which should trigger synthesis) and normal speech variations (which should not). This preliminary filtering prevents false detection while maintaining adaptability to genuine interruptions.
Data Source
AI summary
A method for automatically reading a continuous text composed of several groups of words, as well as a corresponding computer program, storage medium, automatic reader, and user terminal. The method includes providing, in real time, a sound stream corresponding to the text. The sound stream starts from a selected group of words, also called second group of words, selected in the text as a function of at least one result of a real-time analysis of captured speech. The result of the analysis is indicative of a first group of words currently being verbalized by a speaker, the first group of words and the second group of words being different groups of words.