Adaptive Beamforming Filter Update Criterion for Speech Tracking
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing acoustic beamforming algorithms face challenges in accurately tracking speech sources in noisy environments, particularly when sources move or are non-continuous, leading to difficulties in distinguishing desired speech from background noise and adapting to sudden changes.
Innovation Solution
The system employs a beamforming apparatus with adaptive filters and a criterion processor that generates a difference signal to minimize noise interference, allowing for dynamic adjustment of the update criterion based on the normalized difference signal, enabling efficient tracking and acquisition of speech sources even during sudden movements or silence.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If the beamforming algorithm restricts updates to small, slow variations and discards large, sudden variations, then the risk of latching onto noise sources is reduced, but the ability to track large or sudden movements of the desired audio source is restricted
Solution Approach 1:
The patent applies dynamics by making the update criterion adaptive rather than fixed. The beamforming filter's update behavior changes dynamically based on the detected audio signal characteristics. When speech is detected, updates are enabled to track the source; when silence or noise is detected, updates are restricted to prevent locking onto noise sources. This dynamic adjustment resolves the contradiction between reliability and adaptability.
Solution Approach 2:
The patent changes the parameter of the update criterion based on the audio signal state. By detecting whether speech is present and adjusting the update criterion accordingly, the system transitions between different operational modes: a more restrictive mode during silence/noise to ensure reliability, and a more adaptive mode during speech to track source movements. This parameter change resolves the technical contradiction.
2Reliability
If the beamforming algorithm uses a strict update criterion to prevent latching onto noise sources, then reliability is improved, but acquisition performance and ability to lock onto new audio sources deteriorates
Solution Approach 1:
The patent applies preliminary action by detecting the presence of speech before enabling beamforming updates. The system proactively identifies when speech is present and prepares the update criterion accordingly, rather than using a fixed criterion. This preliminary detection allows the system to switch to a more permissive update mode when speech is detected, improving acquisition performance while maintaining reliability during silence or noise conditions.
3Adaptability or versatility
If the beamforming algorithm allows frequent updates to track audio source movements, then tracking performance is improved, but the risk of following interfering noise sources instead of the desired speech source increases
Solution Approach 1:
The patent applies feedback by using the detected audio signal (speech or noise) to control the update criterion. The system continuously monitors the audio input and provides feedback to the beamforming filter's update mechanism. When speech is detected, feedback enables updates to improve tracking; when noise or silence is detected, feedback restricts updates to prevent following noise sources. This feedback loop resolves the contradiction between tracking performance and reliability.
Data Source
AI summary
An apparatus for acoustic beamforming comprises a beamform processor (105) for generating a beamformed signal from two audio inputs. An update processor (107) updates the beamforming filter of the beamform processor (105) if an update criterion is met. An adaptive filter (111) filters the signal from one of the signals and the difference signal between the filtered signal and the signal from the other audio input (101) is generated. An adaptation processor (115) adapts the adaptive filter (111) to minimize the difference signal. A criterion processor (109) modifies the update criterion in response to the (possibly normalized) difference signal. Specifically, the update criterion may be relaxed to improve acquisition performance if the difference signal is indicative of a strong signal outside the beam of the beamform processor (105).