Voice Emotion Parameter Editing for Time-Varying Synthesis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing voice synthesis techniques struggle to set parameters for the transition of time-varying emotions in detail, limiting the ability to create nuanced and dynamic voice expressions.
Innovation Solution
A voice processing support device that allows users to input parameters including multiple types of emotions, mixing ratios, and emotion intensities during voice data reproduction, associating these inputs with specific reproduction timings to generate synthesized voice data with detailed emotional transitions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If a morphing ratio parameter is set for the entire voice data, then voice synthesis is possible, but detailed parameter setting for time-varying emotion transition is difficult
Solution Approach 1:
The patent segments the voice data timeline into multiple time regions, allowing different emotion parameters to be applied to different segments. This enables detailed control of time-varying emotion transitions while maintaining manageable parameter setting complexity through region-based organization
Solution Approach 2:
The patent introduces dynamic emotion parameters that can change over time within different time regions, allowing the emotion mixing ratio to transition smoothly between different emotional states. This creates dynamic voice expressions while the system manages complexity by confining changes to predefined regional boundaries
2Manufacturing precision
If detailed emotion parameters are set for each time point, then time-varying emotion transition precision is improved, but user input load increases
Solution Approach 1:
The patent applies preliminary action by pre-defining time regions and assigning emotion parameters to entire regions rather than individual time points. This approach achieves smooth emotion transitions across time while significantly reducing the number of parameters users must set, thereby reducing input load and time consumption
3Ease of operation
If emotion mixing ratio is fixed for entire voice data, then parameter setting is simple, but dynamic voice expression with changing emotions is limited
Solution Approach 1:
The patent divides the voice data into multiple time regions, each capable of having different emotion mixing ratios. This segmentation allows the system to maintain simplicity within each region while achieving overall dynamic expression flexibility across the entire voice data through the combination of multiple regions
Solution Approach 2:
The patent implements dynamic emotion control by allowing different mixing ratios in different time regions, enabling the voice expression to evolve over time. The system maintains ease of operation by applying static mixing ratios to each region while achieving dynamic overall effect through regional transitions
Data Source
AI summary
According to an embodiment, a voice processing support device includes one or more hardware processors configured to: receive input of a parameter during reproduction of voice data to be edited, the parameter including at least a plurality of types of emotions different from each other and a mixing ratio of the plurality of types of emotions; and record the parameter whose input has been received, in association with a reproduction timing at which the input of the parameter has been received in the voice data.