Speech Synthesizing Dolls Mimicking Parent Voices
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current text-to-speech systems in consumer electronics often use robotic and unfamiliar voices, making it difficult for children to understand audio outputs, particularly when trying to mimic human voices with specific accents and tones.
Innovation Solution
A deep neural network (DNN) is trained using audio recordings of a parent or guardian to replicate their voice, allowing a doll or device to produce audio outputs that mimic the parent's voice while highlighting corresponding text on an electronic display.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If automated text-to-speech systems are used in consumer electronics, then audio output can be provided efficiently, but the voices sound robotic and unfamiliar making it difficult for children to understand
Solution Approach 1:
The patent applies the copying principle by creating voice copies of parents or guardians through deep neural network training. The system records and analyzes the parent's voice characteristics, then generates synthetic speech that closely mimics the original voice. This allows the text-to-speech system to produce familiar, natural-sounding voices that children can easily understand, rather than robotic automated voices.
Solution Approach 2:
The patent utilizes parameter changes by adjusting various acoustic parameters through the deep neural network to transform the synthetic voice into a natural-sounding reproduction of the parent's voice. The system modifies parameters such as pitch, tone, accent, and speech rhythm to match the target voice characteristics, thereby resolving the contradiction between automated efficiency and natural understanding.
2Ease of operation
If deep neural network training is performed to replicate parent voices, then human-like speech with specific accents and tones can be produced, but the device complexity increases
Solution Approach 1:
The patent applies preliminary action by performing the complex deep neural network training process in advance, during an initial setup phase. The system records the parent's voice and trains the DNN model before actual use. Once trained, the model can generate natural-sounding speech without requiring complex real-time processing during normal operation, thus reducing the burden on device complexity during usage.
Solution Approach 2:
The system implements self-service by automatically performing voice recording and DNN training without requiring manual intervention for each speech generation task. The once-off setup process captures the parent's voice characteristics and autonomously creates the voice model, eliminating the need for complex real-time voice synthesis calculations during normal operation.
3Adaptability or versatility
If multiple parents or guardians are supported with separate voice profiles, then the system becomes more versatile, but the configuration process becomes more complex
Solution Approach 1:
The patent applies segmentation by dividing the voice profile management into separate, independent modules for each parent or guardian. Each voice profile is stored and processed as a distinct unit in the system memory, allowing multiple profiles to coexist without interfering with each other. This modular approach simplifies the configuration process, as each profile can be added, modified, or removed independently through simple user interactions.
Data Source
AI summary
Speech synthesizing devices and methods are disclosed for using a doll or toy to mimic the voice of a parent or guardian of a child as reading a children's book. A speech synthesizer that employs an artificial neural network can be used to do so, with the artificial neural network being trained using audio recordings of the parent or guardian speaking as well as text corresponding to the words that are spoken by the parent or guardian in the audio recordings. The artificial neural network may then be used to produce various audio outputs in the voice of the parent or guardian as reading a children's book to the parent or guardian's child.