Multi-Device Hotword Detection With Server-Based Device Selection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In a speech-enabled environment with multiple devices, there is a challenge in ensuring that only one device responds to a hotword command when multiple devices detect the hotword, as users often intend a single device to execute the command.
Innovation Solution
A system where devices communicate with a server to share context data, allowing the server to select the appropriate device to perform speech recognition based on factors like device capabilities, usage history, location, and loudness, while instructing other devices not to process the command.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If multiple devices detect the hotword simultaneously, then the system can respond to user commands from any location, but multiple devices may execute unintended actions
Solution Approach 1:
A server acts as an intermediary between multiple devices and the user command. When devices detect a hotword, they communicate with the server which coordinates the response, ensuring only one device executes the command while others remain idle. This mediator resolves the conflict between multiple devices wanting to respond.
Solution Approach 2:
Devices perform preliminary actions by detecting the hotword and notifying the server before actual command execution. The server then determines which device should execute the command based on context data, preventing multiple devices from simultaneously executing unintended actions.
2Reliability
If all devices process the hotword command, then the system provides comprehensive coverage, but device utilization becomes inefficient
Solution Approach 1:
The command execution function is extracted from all devices and concentrated in a single selected device. While all devices participate in hotword detection, only one device is assigned to execute the actual command, separating the detection function from the execution function.
Solution Approach 2:
Different devices have different roles assigned based on their context. The selected device receives the execution instruction while other devices receive idle instructions, creating local differentiation in device functionality to optimize overall system efficiency.
3Measurement precision
If the system selects a device based on context data, then the intended device responds accurately, but the system complexity increases
Solution Approach 1:
The server serves as a central coordinator that manages the complexity of selecting the appropriate device. It receives context data from multiple devices, processes this information, and makes the selection decision, isolating the complexity from individual devices.
Solution Approach 2:
The system uses context data parameters (device capabilities, usage history, location, loudness) to dynamically determine which device should execute the command. These parameter changes enable accurate device selection without requiring complex hard-coded logic in each device.
Data Source
AI summary
Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for hotword detection on multiple devices are disclosed. In one aspect, a method includes the actions of receiving audio data that corresponds to an utterance. The actions further include determining that the utterance likely includes a particular, predefined hotword. The actions further include transmitting (i) data indicating that the computing device likely received the particular, predefined hotword, (ii) data identifying the computing device, and (iii) data identifying a group of nearby computing devices that includes the computing device. The actions further include receiving an instruction to commence speech recognition processing on the audio data. The actions further include in response to receiving the instruction to commence speech recognition processing on the audio data, processing at least a portion of the audio data using an automated speech recognizer on the computing device.