Home Appliance Voice Recognition Noise Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing home appliances that generate large magnitude noise pose challenges for voice recognition systems, as conventional noise removal methods struggle to effectively distinguish voice from noise when the noise level exceeds that of the user's voice.
Innovation Solution
A home appliance configuration that includes a first microphone on the surface and a second microphone inside the housing, along with a processor that generates noise data from the second voice data and uses it for signal processing of the first voice data to improve voice recognition accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conventional noise removal methods are used in home appliances that generate large magnitude noise, then the device can operate normally, but the voice recognition rate deteriorates because the noise level exceeds the user's voice level
Solution Approach 1:
The patent divides the noise removal task into two segments: first, a rough noise removal stage that eliminates obvious background noise, and second, a precise noise removal stage that specifically targets appliance-generated noise. This segmentation allows the system to handle different noise characteristics at different processing stages, improving overall voice recognition accuracy without being overwhelmed by the total noise level.
Solution Approach 2:
The patent applies preliminary noise removal processing before voice recognition. By pre-processing the audio signal to reduce noise levels and enhance voice components beforehand, the system prepares the input data in a more favorable state for the subsequent recognition stage, thereby improving the voice recognition rate in high-noise environments.
2Power
If the home appliance generates large magnitude noise for its operation, then the appliance can perform its function effectively, but it becomes difficult to distinguish user voice from the generated noise
Solution Approach 1:
The patent introduces an intermediary noise removal processing stage between the noisy audio input and the voice recognition system. This intermediary process analyzes the audio signal, identifies appliance-generated noise patterns, and selectively removes them while preserving the user's voice. This mediator enables the system to maintain high-power appliance operation while achieving accurate voice detection despite the presence of strong operational noise.
Data Source
AI summary
A home appliance including a first microphone that is disposed on a surface of a housing, a second microphone that is disposed on an inside of the housing, and a processor configured to perform signal processing for first voice data that is acquired from the first microphone, and perform voice recognition using the signal-processed first voice data. The processor is further configured to generate noise data using second voice data that is acquired from the second microphone and perform signal processing for the first voice data using the generated noise data.