Connected Speaker AI Sharing for Lower-Cost Multi-Service Audio
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The increasing diversity of AI speakers with different AI services in homes leads to challenges such as high hardware and software costs, space occupation, and inter-device interference, making it difficult to satisfy customized user demands efficiently.
Innovation Solution
A data processing method that enables sharing of microphone data and AI audio data between multiple speakers through a data connection, determining an awakened AI speaker based on captured microphone data, and generating AI audio data using an AI functional module to share AI functionalities across speakers.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If multiple speakers with different AI services are deployed to satisfy customized user demands, then the functionality and adaptability of the speaker system is improved, but the device complexity and cost increase
Solution Approach 1:
The patent merges multiple speakers into a collaborative system where they share AI processing capabilities. Instead of each speaker independently handling AI tasks, the system combines their microphones and processing power, allowing simpler speakers to participate in AI conversations by sharing resources with AI-capable speakers.
Solution Approach 2:
The patent enables speakers without AI functional modules to perform AI-related tasks by collaborating with AI-capable speakers. Non-AI speakers can participate in AI conversations, wake-word detection, and voice assistant interactions through the shared system, making them multi-functional beyond their original capabilities.
2Adaptability or versatility
If multiple AI speakers are deployed to provide diverse AI services, then the adaptability is improved, but the hardware cost and space occupation increase
Solution Approach 1:
The patent combines hardware resources across multiple speakers, allowing AI processing to be shared. This reduces the need for every speaker to have full AI capabilities locally, as AI-capable speakers can serve the entire group, thereby reducing overall hardware requirements.
Solution Approach 2:
The patent allows non-AI speakers to effectively 'copy' AI capabilities by accessing and utilizing the AI processing power of AI-capable speakers through the shared system, without needing to duplicate the expensive AI hardware in each device.
3Reliability
If each speaker has its own AI functional module to ensure independent operation, then the reliability is improved, but the device complexity and cost increase
Solution Approach 1:
The patent introduces a communication bus as an intermediary that connects speakers with AI capabilities to non-AI speakers. This mediator enables resource sharing and coordinated operation, allowing speakers to maintain reliability through the networked system rather than requiring independent AI modules in each device.
4Device complexity
If speakers share AI processing capabilities through a data connection, then the device complexity is reduced, but the measurement precision of individual speaker performance may deteriorate
Solution Approach 1:
The patent segments the AI processing function from the audio capture function. Microphones on all speakers continue to capture local audio independently with full precision, while AI processing is selectively performed by AI-capable speakers. This segmentation maintains measurement precision for audio input while reducing overall system complexity.
Data Source
AI summary
A data processing method for a plurality of speakers connected to each other is provided. The plurality of speakers includes at least one artificial intelligence (AI) speaker integrated with an AI module and having a microphone and at least one non-AI speaker integrated with no AI module and having a microphone. The method includes capturing microphone data through the plurality of speakers, wherein the speaker from which the captured microphone data originates is a source speaker, determining an awakened AI speaker of the plurality of speakers based on the captured microphone data, generating AI audio data for the captured microphone data using an AI module in the awakened AI speaker, and playing the generated AI audio data using at least the source speaker.