Distributed Speech Synthesis Using Segment Indexing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing text-to-speech synthesis systems require high memory and computational power in end terminals and result in large data transmission volumes, especially when used in distributed environments like mobile devices.
Innovation Solution
A method that determines and transmits only the necessary speech segments to the terminal, using indexing to reduce memory and computational requirements, with a cache for reusing segments and a server managing the index lists to minimize data transfer.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If TTS is implemented centrally on a server, then speech quality can be maintained, but data volume to be transmitted becomes relatively high (>4.8 kbit/s)
Solution Approach 1:
The patent divides the speech synthesis database into discrete speech segments (phonemes, diphones, words, or phrases) that can be individually stored, transmitted, and concatenated. This segmentation allows the system to transmit only necessary segments rather than entire speech databases, reducing data volume while maintaining quality through precise segment selection and concatenation.
Solution Approach 2:
The system pre-phonetizes text and prepares speech segments in advance on the server side. By pre-processing and preparing segments before transmission, the system avoids the need to transmit large amounts of raw data and enables efficient assembly at the terminal, reducing transmission requirements while maintaining speech quality.
2Quantity of substance
If TTS is implemented in the end terminal, then data transmission is minimized, but memory requirements become large and computational power requirements increase
Solution Approach 1:
By segmenting the speech database into smaller units (phonemes, diphones, words, phrases), the system reduces the memory footprint at terminals. Instead of storing entire speech databases, terminals only need to store necessary segments and their indices, significantly reducing memory and computational requirements while enabling distributed processing.
Solution Approach 2:
The patent introduces an indexing mechanism as an intermediary between the speech segments and the terminal. Indices serve as references that allow terminals to access and assemble segments without needing to store the complete database, reducing local resource requirements while maintaining the ability to generate high-quality speech.
3Reliability
If all speech segments are stored in the terminal database, then speech quality is ensured, but memory capacity requirements become large
Solution Approach 1:
The patent segments the speech database into discrete, manageable units that can be selectively stored and transmitted. This segmentation enables the system to maintain speech quality by using appropriate segments while reducing memory capacity requirements by storing only necessary segments rather than entire databases.
Solution Approach 2:
The system implements local quality by storing and using only the specific speech segments needed for particular speech outputs rather than maintaining a complete database. This allows terminals to achieve high speech quality for specific tasks while using minimal memory capacity by loading only relevant segments on demand.
Data Source
AI summary
In a method of generating speech from text the speech segments necessary to put together the text to be output as speech by a terminal are determined; it is checked, which speech segments are already present in the terminal and which ones need to be transmitted from a server to the terminal; the segments to be transmitted to the terminal are indexed; the speech segments and the indices of segments to be output at the terminal are transmitted; an index sequence of speech segments to be put together to form the speech to be output is transmitted; and the segments are concatenated according to the index sequence. This method allows to realize a distributed speech synthesis system requiring only a low transmission capacity, a small memory and low computational power in the terminal.