Contextual Beamforming for Noisy Audio Input
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Legacy beamforming algorithms in information handling systems are ineffective in noisy environments, particularly when users are not directly in front of the microphone array, and adaptive beamforming may not accurately center on the correct user, leading to poor signal-to-noise ratio (SNR) in audio input processing.
Innovation Solution
An audio processing system incorporating a microphone array, a speech detection system, and a neural network noise reduction module that includes a speaker recognition module, delay calculation module, and adaptive beamforming module to enhance audio signals from recognized users while reducing noise, even in noisy environments.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If legacy beamforming algorithms are used, then device complexity is reduced, but signal-to-noise ratio deteriorates in noisy environments
Solution Approach 1:
The audio processing system is segmented into distinct functional modules: speech detection system, beamforming processing module, and neural network noise reduction module. Each module performs a specific function in sequence, allowing complex noise reduction to be achieved through modular processing stages rather than a single complex algorithm.
Solution Approach 2:
The beamforming processing acts as an intermediary between the raw audio signal and the neural network noise reduction module. It pre-processes the audio signal by enhancing speech from specific directions, which improves the input quality for the subsequent neural network noise reduction stage.
2Reliability
If adaptive beamforming is used to track user location, then audio signal enhancement improves, but accuracy deteriorates when users are not directly in front of the microphone array
Solution Approach 1:
The beamforming processing is made dynamic and adaptive, automatically adjusting to track the location of speaking users in real-time. The system dynamically updates beamforming parameters based on detected speech sources, allowing it to follow users as they move around the environment rather than being fixed to a predetermined position.
Solution Approach 2:
The system changes beamforming parameters (such as beam direction and width) based on detected speech source locations. When a user speaks, the system identifies their direction and adjusts the beamforming parameters to enhance audio from that specific direction, adapting to different user positions throughout the environment.
3Reliability
If neural network noise reduction is applied to all audio input, then noise reduction performance improves, but processing efficiency deteriorates
Solution Approach 1:
Instead of applying neural network noise reduction to all audio input continuously, the system applies it selectively based on speech detection. The neural network processes audio segments when speech is detected, rather than continuously processing all audio, achieving effective noise reduction while reducing overall processing load.
Solution Approach 2:
The beamforming processing is performed as a preliminary action before neural network noise reduction. By pre-enhancing the audio signal with beamforming, the neural network receives higher quality input, which improves its noise reduction effectiveness and allows it to process more efficiently with better results.
Data Source
AI summary
An audio processing system includes a microphone array, a speech detection system, and a neural network noise reduction module. The microphone array includes at least two microphones and provides an audio signal from an environment surrounding the microphone array. The speech detection system receives the audio signal, and processes the audio signal to a) detect that a first user is speaking, b) determine a first direction relative to the audio array when the first user is located at a first location within the environment, and c) provide beamforming processing on the audio signal in the first direction, and to provide a processed audio signal based upon the beamforming processing. The neural network noise reduction module reduces noise in the processed audio signal.