Dynamic Speech Synthesis Path Selection for Load Adaptation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing speech synthesis systems are inflexible and cannot adapt to varying load conditions, leading to potential overload and delayed user feedback, especially during high demand periods.
Innovation Solution
A method and device that dynamically select a speech synthesis path based on the system's load level, using different models such as LSTM, HTS, and vocoder models to optimize performance and ensure timely service.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If the speech synthesis system uses a fixed configuration, then the system structure is simple and easy to maintain, but the system cannot adapt to varying load conditions and may overload during high demand periods
Solution Approach 1:
The speech synthesis system dynamically selects between different synthesis paths (first path with first model, second path with second model) based on real-time load levels. The system transitions from a fixed configuration to a dynamic one where the processing path is adjusted according to current operational conditions, enabling adaptation to varying demand without increasing overall system complexity.
Solution Approach 2:
The system changes the parameter of model selection based on load level parameters. When load level exceeds a threshold, the system switches from using the first model to the second model, effectively changing the operational parameters to maintain performance under different conditions without requiring a complete system redesign.
2Productivity
If the speech synthesis system processes a large number of requests simultaneously, then the productivity increases, but the system may become overloaded and cause delays in user feedback
Solution Approach 1:
The speech synthesis system is segmented into multiple processing paths, each handling different load conditions. The first path processes requests when load is below the threshold, while the second path handles requests when load exceeds the threshold. This segmentation allows the system to maintain high productivity by distributing requests across different paths while ensuring reliability by preventing any single path from becoming overloaded.
Solution Approach 2:
The load level determination mechanism acts as an intermediary that mediates between incoming requests and the appropriate synthesis path. This intermediary monitors system state and directs requests to suitable processing paths, ensuring that productivity is maintained while preventing overload conditions that would compromise service stability.
3Adaptability or versatility
If the system uses a single speech synthesis model, then the device complexity is reduced, but the system cannot flexibly adapt to different load levels and user experience deteriorates during peak usage
Solution Approach 1:
The speech synthesis system achieves multi-functionality by implementing multiple synthesis paths that can handle different operational scenarios. The first path serves normal load conditions while the second path serves high load conditions, making the system universal enough to adapt to various load levels without requiring completely separate systems for each scenario.
Data Source
AI summary
The present invention provides a method and a device for optimizing speech synthesis system. The method comprises: receiving speech synthesis requests contained text messages; and determining the load level of the speech synthesis system when the speech synthesis requests are received; and selecting speech synthesis paths corresponding to the load level and synthesizing the text into speech according to the speech synthesis paths.