Probabilistic Beamforming for Accurate Target Speech Extraction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing speech recognition systems struggle to accurately separate target speech from noise in input signals, leading to interference and reduced performance.
Innovation Solution
A beamforming device that estimates a speech existence probability based on an input vector, using a Bayesian approach to calculate a steering vector and weight vector, thereby enhancing the extraction of target speech signals.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional beamforming methods are used to extract target speech signals, then the device structure remains simple, but the speech extraction accuracy deteriorates due to noise interference
Solution Approach 1:
The patent applies preliminary action by estimating speech existence probability before performing beamforming weight calculation. The probability estimation unit processes input signals to determine the likelihood of speech presence in each signal source, and this probability information is then used by the beamforming unit to adjust weight vectors. This preliminary probability estimation enables more accurate speech extraction while maintaining a relatively simple device structure.
2Measurement precision
If speech existence probability estimation is performed to improve speech extraction accuracy, then the extraction precision improves, but the processing complexity increases
Solution Approach 1:
The patent applies parameter changes by transforming the beamforming approach from deterministic weight calculation to probabilistic weight adjustment. Instead of using fixed beamforming weights, the system calculates weight vectors based on speech existence probability values that range from 0 to 1. This parameter transformation allows the system to adaptively adjust beamforming weights according to the estimated probability of speech presence, improving extraction accuracy while keeping the processing complexity manageable through efficient probability estimation algorithms.
Data Source
Figure 1~2
Figure 3~5
Figure 6
AI summary
A beamforming device according to an embodiment of the present invention may include a probability estimation unit, a steering vector unit, and a beamforming unit. The probability estimation unit may estimate a speech existence probability corresponding to a probability that a target speech signal exists based on an input vector. The steering vector unit may provide an estimated steering vector according to the speech existence probability and an input vector. The beamforming unit may calculate a weight vector based on the speech existence probability, the input vector, and the estimated steering vector to provide an output vector. According to the beamforming device of the present invention, it is possible to more accurately extract the target speech signal from the input signal by estimating the speech existence probability corresponding to the probability that the target speech signal exists based on the input vector to provide the steering vector and the weight vector.