Local Voice Query Interpretation with Selective Remote Fallback
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current voice control systems require network connectivity and significant bandwidth to transmit voice commands to remote servers for transcription due to the complexity of voice queries, leading to increased costs and data transmission demands.
Innovation Solution
A system that locally interprets voice queries by comparing audio characteristics with stored queries, transmitting only unmatched queries to a remote server, and managing storage size by removing infrequently used or old data to reduce bandwidth and storage needs.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If voice commands are transmitted to a remote server for transcription, then recognition accuracy is improved, but data transmission bandwidth requirements increase
Solution Approach 1:
The patent segments voice queries into two categories: common queries handled locally with pre-stored transcriptions, and uncommon queries transmitted to remote servers. This segmentation allows the system to process the majority of queries locally without bandwidth consumption while maintaining high recognition accuracy for common commands, thereby resolving the contradiction between recognition accuracy and bandwidth usage.
Solution Approach 2:
The system performs preliminary action by pre-storing transcriptions and audio characteristics of common voice queries locally on the device. This preliminary preparation enables immediate local processing of frequent queries without requiring real-time network transmission, thus improving recognition speed and reducing bandwidth requirements while maintaining accuracy for common commands.
2Productivity
If local storage of voice query data is increased, then recognition speed is improved, but device storage capacity is exceeded
Solution Approach 1:
The patent applies parameter changes by storing only essential audio characteristics (duration, tone, rhythm, cadence, accent) rather than complete voice query recordings. This parameter-based approach significantly reduces the storage footprint while maintaining sufficient information for accurate local matching, thereby improving recognition speed without exceeding device storage capacity.
Solution Approach 2:
The system extracts only the necessary audio characteristics from complete voice queries for local storage. By taking out and storing only the essential parameters needed for matching (duration, tone, rhythm, cadence, accent) rather than the full audio data, the system achieves fast local recognition while keeping storage requirements within device limits.
3Quantity of substance
If more voice commands are recognized locally, then network bandwidth usage is reduced, but device complexity increases
Solution Approach 1:
The patent uses copying by storing simplified representations (audio characteristics and transcriptions) of common voice queries locally. This copying approach allows the device to perform simple local matching operations without implementing complex full speech-to-text processing capabilities, thereby reducing network bandwidth usage while keeping device complexity manageable through the use of pre-prepared local copies.
Data Source
AI summary
Systems and methods are described herein for locally interpreting a voice query and for managing a storage size of data stored locally to support such local interpretation of voice queries. A voice query is received and compared with a plurality of stored voice queries having similar audio characteristics. If a match is identified, text corresponding to the matching stored voice query is retrieved, and an action corresponding to the retrieved text is performed. If the locally stored table does not contain a stored voice query that matches the voice query, the voice query is transmitted to a remote server for transcription. Once the transcription is received from the remote server, the voice query and the transcription are stored in the table in association with one another.