Systems and methods for classifying sounds
AssigneeApple Inc.
InventorsDaniel C. Klingler, Carlos M. Avendano, Hyung-Suk Kim, Miquel Espi Marques
Life of this patent
Estimating the term…
Abstract
An electronic device has one or more microphones that pick up a sound. At least one feature extractor processes the audio signals from the microphones, that contain the picked up the sound, to determine several features for the sound. The electronic device also includes a classifier that has a machine learning model which is configured to determine a sound classification, such as artificial versus natural for the sound, based upon at least one of the determined features. Other aspects are also described and claimed.
Claims20
An electronic device, comprising:
- one or more microphones configured to receive a sound;
- a processor and memory having stored therein a plurality of instructions that when executed by the processor implement
- at least one feature detector configured to receive one or more audio signals from the microphones that contain the sound, process the audio signals to determine a plurality of features for the audio signals including a directional feature and a sound class feature;
- and
- a sound classifier including a machine learning model that is configured to receive the determined features from the at least one feature detector, and determine a sound classification for the audio signal based upon at least one of the determined features.
An electronic device, comprising:
- a plurality of microphones configured to produce a plurality of audio signals in response to receiving a sound;
- at least one feature detector configured to receive the plurality of audio signals from the plurality of microphones, and determine a plurality of features relating to the sound including a directional feature, a sound class feature, and a distortion feature;
- and
- a classifier including a machine learning model that is configured to receive the determined features, and determine a sound classification for the sound based upon at least one of the determined features,
- wherein the electronic device performs one or more functions or actions based upon the determined sound classification.
A method performed by a processor of an electronic device for discriminating between two classes of sounds, comprising:
- capturing a sound using a plurality of microphones, as a recorded sound;
- digitally processing the recorded sound to determine at least two features of the recorded sound that include a directional feature and a sound class feature;
- determining a classification of the sound based on the determined features using a machine learning model;
- and
- performing a virtual assistant action based upon the determined classification.
Dependencies are inferred from the claim text.