Streaming Voice Conversion Segmentation for Real-Time Processing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing voice conversion systems face challenges in achieving real-time conversion while maintaining fidelity, as they require waiting for the user to finish speaking before processing can begin, leading to difficulties in handling long-time voice input.
Innovation Solution
A streaming voice conversion method that partitions voice data into smaller segments, processes each segment in parallel using a multi-core processor, and restores them for seamless output, allowing for real-time conversion and improved speed without compromising fidelity.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If the existing voice conversion system waits for the user to finish speaking before processing, then the conversion fidelity is maintained, but the conversion speed and real-time performance deteriorate
Solution Approach 1:
The patent divides the voice conversion process into multiple partitions or segments. Each partition processes a portion of the voice data independently and in parallel, allowing the system to start processing before the user finishes speaking. This segmentation enables real-time conversion while maintaining fidelity by processing multiple segments simultaneously through a multi-core processor.
2Productivity
If the system processes long-time voice input in real-time, then the conversion speed improves, but the difficulty of maintaining fidelity increases
Solution Approach 1:
The patent performs preliminary actions by preparing and processing voice data partitions in advance and in parallel before the complete voice input is received. The system pre-processes multiple partitions simultaneously using a multi-core processor, so that when the voice input is complete, the conversions are already prepared or nearly prepared, enabling real-time output while maintaining quality through careful partition management and restoration.
3Measurement precision
If the voice data is processed as a single unit, then the fidelity is preserved, but the response time and processing efficiency deteriorate
Solution Approach 1:
The patent segments the voice data into multiple partitions that can be processed independently and in parallel. Each partition is handled as a separate unit by different processor cores, significantly reducing the overall processing time compared to sequential processing of a single unit. The partitions are then restored and combined to maintain the original voice quality and fidelity.
Solution Approach 2:
The patent merges the results from multiple parallel partition processing operations to produce the final converted voice output. By combining the individually processed partitions through a restoration process, the system achieves both fast response time from parallel processing and high fidelity from quality-preserving conversion algorithms.
Data Source
AI summary
The present disclosure provides a streaming voice conversion method as well as an apparatus and a computer readable storage medium using the same. The method includes: obtaining to-be-converted voice data; partitioning the to-be-converted voice data in an order of data obtaining time as a plurality of to-be-converted partition voices, where the to-be-converted partition voice data carries a partition mark; performing a voice conversion on each of the to-be-converted partition voices to obtain a converted partition voice, where the converted partition voice carries a partition mark; performing a partition restoration on each of the converted partition voices to obtain a restored partition voice, where the restored partition voice carries a partition mark; and outputting each of the restored partition voices according to the partition mark carried by the restored partition voice. In this manner, the response time is shortened, and the conversion speed is improved.