Multi-Microphone Voice Separation Using Phase-Reversed Noise Canceling
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Robots face challenges in accurately recognizing and separating voice signals from multiple users in a vehicle environment, leading to errors in service provision when multiple users request services simultaneously.
Innovation Solution
A method and system utilizing multiple microphones and a processor to reverse the phase of sound signals from other microphones, generating noise-canceling signals to extract individual voice signals, allowing for accurate separation and recognition of voice commands from multiple users.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If multiple microphones are used to receive voice signals from multiple users, then the ability to capture multiple voice signals improves, but the difficulty of separating and accurately recognizing individual voice signals increases
Solution Approach 1:
The patent divides the mixed sound signals into individual voice signals by separating them according to their respective sources. Each microphone's received signal is processed independently, and voice signals from different users are segmented through phase reversal and noise canceling operations to extract individual voices from the mixture.
Solution Approach 2:
The patent introduces phase-reversed sound signals from other microphones as intermediary elements to cancel out interfering voices. By using these intermediary signals with reversed phases, the system能够有效ly removes noise and interference from target voice signals, enabling accurate extraction of individual voices.
2Measurement precision
If phase reversal and noise canceling operations are performed to separate voice signals, then voice recognition accuracy improves, but the processing complexity and computational requirements increase
Solution Approach 1:
The patent performs preliminary phase reversal operations on sound signals from other microphones before combining them with the target microphone's signal. This preliminary action of reversing phases in advance simplifies the subsequent noise canceling process and makes the overall signal separation more manageable and efficient.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
Enables effective conversation and service provision without errors even when multiple users output voice signals simultaneously, improving the robot's ability to understand and respond to multiple requests in real-time.
Implementation Method 1
the processor reverses a phase of at least one of other sound signals received respectively by at least one of other microphones except an i-th microphone among the plurality of microphones
Implementation Method 2
generating noise-canceling signals to extract individual voice signals accurately
Data Source
AI summary
Disclosed herein are a method for extracting voice signals of a plurality of users, and a terminal device and a robot implementing the same. The robot includes a plurality of microphones respectively corresponding to a plurality of users, a memory and a processor, where the processor extracts a plurality of voice signals output by each of the plurality of users using a plurality of sound signals received by each of the plurality of microphones, and in this case, the processor reverses a phase of at least one of other sound signals received respectively by at least one of other microphones except an i-th microphone, and extracts an i-th voice signal output by an i-th user corresponding to the i-th microphone based on a i-th sound signal received by the i-th microphone and based on at least one of other sound signals, the phase of which is reversed.