Self-Service Terminal Voice Filtering for Same-Direction Noise
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional beam-forming mechanisms in self-service registration terminals struggle to effectively distinguish between useful speech signals and interference noise in noisy environments, particularly when sound sources are located in the same direction, leading to impaired speech recognition.
Innovation Solution
A self-service terminal employs a mechanism that filters sound based on transit time differences and sound pressure differences between multiple microphones to separate useful speech signals from interference noise, using a minimum and maximum boundary in depth to enhance signal reception.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Length of moving object
If conventional beam-forming mechanisms are used to focus sound reception in a specific direction, then lateral directivity is improved, but signals from sound sources located one behind the other (same direction) cannot be effectively distinguished
Solution Approach 1:
The patent extends beam-forming from two-dimensional lateral directionality to three-dimensional spatial filtering by incorporating depth information through transit time differences. Multiple microphones arranged in space create independent transit time differences for sound sources at different depths, enabling the system to distinguish between front and back sources that project to the same lateral position.
Solution Approach 2:
The patent introduces transit time difference as an intermediary parameter to mediate between the microphone array and sound source localization. By calculating and comparing transit time differences across multiple microphone pairs, the system creates an intermediate representation that encodes depth information, which then feeds into the speech recognition process.
2Measurement precision
If multiple microphones are used to improve speech recognition in noisy environments, then signal detection capability is improved, but the complexity of signal processing increases
Solution Approach 1:
The patent segments the complex multidimensional signal processing task into sequential stages: first calculating transit time differences for each microphone pair, then using these differences to determine spatial characteristics, and finally applying appropriate filtering. This segmentation breaks down the complex problem into manageable computational steps.
Solution Approach 2:
The patent performs preliminary calculation of transit time differences between all microphone pairs before the actual speech recognition process. These pre-calculated time differences are stored and used to configure the beam-forming filters in advance, avoiding real-time complex calculations during speech processing.
3Power
If beam-forming is applied to amplify signals from a specific direction, then useful signal strength is improved, but interference signals from the same direction are also amplified
Solution Approach 1:
The patent applies different filtering characteristics to different spatial regions by using transit time difference analysis. Instead of uniform amplification in a directional beam, the system creates localized enhancement zones at specific depths and positions, amplifying only the desired speech source while leaving other regions unaffected.
Solution Approach 2:
The patent converts the harmful effect of interference signals into a useful diagnostic tool. By analyzing how interference signals affect transit time differences across microphone pairs, the system identifies and excludes these signals, transforming what would be noise into information about the acoustic environment that aids in speech source identification.
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
This approach improves speech recognition accuracy by distinguishing between desired and undesired voice signals, even in noisy conditions, by leveraging transit time and sound pressure differences to filter out interference, thus enhancing the functionality of self-service terminals.
Implementation Method 1
determine an origin location of sound sources in a three-dimensional space on the basis of transit time differences between the acoustic sensors
Implementation Method 2
a plurality of acoustic sensors; and a control device configured to: superimposing a signal detected by the plurality of acoustic sensors
Data Source
AI summary
A self-service terminal (100) can have:—a product-sensing device for sensing a property of a product;—a plurality of acoustic sensors (104a, 104b); and a control device, which is designed for: superposing a signal captured by means of the plurality of acoustic sensors (104a, 104b); determining a voice pattern on the basis of the result of the superposing; outputting information on the basis of the property and on the basis of the voice pattern; wherein the superposing and the position of the plurality of acoustic sensors (104a, 104b) relative to each other are designed such that first components of the signal are attenuated relative to second components of the signal if an origin of the second components is located between the self-service terminal (100) and an origin of the first components.