In-Car Multi-Passenger Voice Dialog With Seat-Specific Speech Processing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In a passenger vehicle, acoustic cross-coupling between microphones hinders effective communication between passengers and an automotive assistant, preventing other passengers from contributing to or taking over a dialog initiated by one passenger.
Innovation Solution
Implementing seat-specific speech daemons with automatic speech-recognition and natural-language understanding units to analyze utterances from each seat, allowing the automotive assistant to distinguish between relevant and irrelevant contributions, and manage concurrent dialogs.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If the automotive assistant directs full attention to one passenger using signal processing methods to isolate speech signals, then dialog robustness is improved and interfering signals are reduced, but other passengers cannot contribute to or take over the dialog
Solution Approach 1:
The system segments the dialog management by creating seat-specific speech daemons, where each microphone/acoustic zone has its own dedicated speech processing instance. This segmentation allows the system to maintain isolated speech signals for reliability while also enabling multiple passengers to independently initiate and contribute to dialogs, resolving the contradiction between focusing on one passenger and allowing multi-passenger participation.
2Measurement precision
If microphones are placed close to each seat for clear speech capture, then speech signal quality is improved, but acoustic cross-coupling between microphones increases and hinders communication
Solution Approach 1:
The system extracts and removes acoustic cross-coupling from the microphone signals using signal processing methods. By isolating the harmful cross-coupling component and separating it from the desired speech signals, the system maintains close microphone placement for good signal quality while eliminating the interfering acoustic coupling between adjacent microphones.
3Reliability
If the automotive assistant ignores stray utterances from other passengers during a dialog, then interfering signals are reduced and dialog robustness increases, but relevant contributions from other passengers are missed
Solution Approach 1:
The system implements feedback mechanisms where the dialog state and active speaker information are continuously monitored and used to adjust speech processing in real-time. This feedback allows the system to distinguish between irrelevant stray utterances that should be ignored and relevant contributions from other passengers that should be incorporated, maintaining dialog robustness while capturing useful information from multiple sources.
Data Source
AI summary
An automotive assistant that is connected to microphones and loudspeakers that are associated with different seats in a passenger vehicle includes a dialog manager that is configured to initiate a dialog based on an utterance received at a first one of the microphones and to advance that dialog based on an utterance received from another of the microphones.