Intel Corporation (20240223948). MICROPHONE CHANNEL SELF-NOISE SILENCING simplified abstract
Contents
MICROPHONE CHANNEL SELF-NOISE SILENCING
Organization Name
Inventor(s)
Adam Kupryjanow of Gdansk (PL)
Przemyslaw Maziewski of Gdansk (PL)
Lukasz Pindor of Pruszcz Gdanski (PL)
Sebastian Rosenkiewicz of Gdansk (PL)
MICROPHONE CHANNEL SELF-NOISE SILENCING - A simplified explanation of the abstract
This abstract first appeared for US patent application 20240223948 titled 'MICROPHONE CHANNEL SELF-NOISE SILENCING
The patent application describes a user computing device with a microphone that generates an audio signal and a self-noise silencer that generates a feature set corresponding to the audio signal. The feature set includes magnitude values for different frequency components in the audio signal. A machine learning model is used to infer frequencies contributing to self-noise generated at the microphone, and an attenuation mask is generated based on the model's output to identify attenuation values for certain frequency components. This attenuation mask is then applied to the magnitude values to remove self-noise and produce a denoised version of the audio signal.
- User computing device with microphone and self-noise silencer
- Feature set includes magnitude values for frequency components in audio signal
- Machine learning model used to infer frequencies contributing to self-noise
- Attenuation mask generated to identify attenuation values for specific frequency components
- Mask applied to magnitude values to remove self-noise and generate denoised audio signal
Potential Applications: - Noise cancellation in audio recording devices - Improving voice recognition accuracy in noisy environments - Enhancing audio quality in communication devices
Problems Solved: - Reducing self-noise in audio signals - Improving overall audio signal quality - Enhancing user experience with audio devices
Benefits: - Clearer audio recordings - Improved accuracy in voice recognition - Better communication in noisy environments
Commercial Applications: Title: "Advanced Noise Cancellation Technology for Audio Devices" This technology can be used in smartphones, laptops, and other audio recording devices to provide users with high-quality audio recordings and clear communication in various environments. The market implications include increased demand for devices with superior noise cancellation features, leading to potential partnerships with audio device manufacturers.
Prior Art: Readers can start their search for prior art related to this technology by looking into patents or research papers on noise cancellation techniques in audio devices, machine learning models for audio signal processing, and self-noise reduction methods in microphones.
Frequently Updated Research: Researchers are constantly exploring new machine learning algorithms for audio signal processing, advancements in noise cancellation technology, and innovative approaches to improving audio quality in various devices.
Questions about Noise Cancellation Technology: 1. How does this technology compare to traditional noise cancellation methods? Traditional noise cancellation methods typically rely on simple filters or algorithms to reduce background noise, while this technology uses machine learning to identify and remove self-noise specifically generated at the microphone, resulting in more precise noise cancellation. 2. What are the potential limitations of this technology in real-world applications? The effectiveness of this technology may vary depending on the complexity of the audio environment and the quality of the microphone. Additionally, the computational resources required for running the machine learning model could be a limiting factor in certain devices or applications.
Original Abstract Submitted
a user computing device includes a microphone to generate an audio signal and a self-noise silencer to generate a feature set corresponding to the audio signal, where the input feature identifies, for each of a plurality of frequency components in the audio signal, a respective magnitude value. at least a portion of the feature set is provided as an input to a machine learning model trained to infer frequencies contributing to self-noise generated at the microphone. an attenuation mask is generated, based on an output of the machine learning model, that identifies an attenuation value for at least a subset of the plurality of frequency components. the attenuation mask is applied to at least the subset of the magnitude values of the plurality of frequency components to remove self-noise from the audio signal and generate a denoised version of the audio signal.
