Dynamic TTS Dictionary Update for Broadcast Text
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Text-to-speech (TTS) systems in broadcast receivers face challenges in handling unlimited vocabularies, particularly with transient data like EPG and teletext, where unknown words or abbreviations cannot be predicted in advance, leading to difficulties in pronunciation for blind or partially sighted users.
Innovation Solution
The system updates the dictionary data used by TTS systems with broadcast updates, allowing for dynamic addition of pronunciations for unknown words through a network connection or remote repository, enabling users to request corrections and incorporating updates into the conversion dictionary and rules database.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If TTS systems use pre-recorded voices or phoneme-based conversion with fixed lexicons, then pronunciation accuracy for known words is improved, but the system cannot handle unknown words or abbreviations in transient data like EPG and teletext
Solution Approach 1:
The patent implements a dynamic dictionary update mechanism where the TTS system can receive and incorporate new word pronunciations broadcast via the broadcast signal. This allows the fixed lexicon to become dynamically expandable, resolving the contradiction between maintaining pronunciation accuracy for known words while adapting to unknown words in transient data.
Solution Approach 2:
The system enables self-updating of the pronunciation dictionary through automated reception of broadcast updates containing new vocabulary. The TTS system serves itself by automatically incorporating new words without requiring manual intervention, thus expanding vocabulary coverage while maintaining accuracy.
2Adaptability or versatility
If the TTS dictionary is updated manually by users to add new words, then vocabulary coverage is improved, but this places an undue burden on blind or partially sighted users and is impractical for transient data
Solution Approach 1:
The system automatically updates its dictionary through broadcast signals without requiring user intervention. The receiver passively receives dictionary updates embedded in the broadcast stream, eliminating the need for blind or partially sighted users to manually input or correct pronunciation data.
Solution Approach 2:
The broadcast signal serves as an intermediary that delivers dictionary updates from the content provider directly to the TTS system. This mediator mechanism transfers vocabulary updates without requiring direct user interaction, thus improving ease of operation while expanding vocabulary coverage.
3Device complexity
If the TTS system uses a fixed dictionary compiled in advance, then system complexity is reduced and pronunciation accuracy for known words is maintained, but the system cannot adapt to new abbreviations or names that appear in broadcast content
Solution Approach 1:
The patent transforms the static dictionary into a dynamic structure that can be updated through broadcast signals. The TTS system maintains a base fixed dictionary for simplicity while incorporating a dynamic update mechanism that receives new vocabulary via broadcast, thus balancing low complexity with high adaptability.
Solution Approach 2:
The system performs preliminary compilation of a base dictionary in advance to maintain simplicity, while preparing for future updates through the broadcast mechanism. This preliminary action establishes the core functionality with low complexity while enabling future adaptability through automated updates.
Data Source
Figure 1~3
Figure 2~4
Figure 5
AI summary
A broadcast signal receiver comprises a text data receiver for receiving broadcast text data for display to a user in relation to a user interface; a text-to-speech (TTS) converter for converting received text data into an audio speech signal, the TTS converter being operable to detect whether a word for conversion is included in a stored list of words for conversion and, if so, to convert that word according to a conversion defined by the stored list; and if not, to convert that word according to a set of predetermined conversion rules; a conversion memory storing the list of words for conversion by the TTS converter; and an update receiver for receiving additional words and associated conversions for storage in the conversion memory.