# US 12,002,442 B2 · Method and device for personalized sound masking in vehicle > Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt - Number: US 12,002,442 B2 - Document ID: US12002442B2 - Status: Granted - Kind: B2 - Application number: 17/878,586 - Filed: 2022-08-01 - Published: 2023-06-08 - Granted: 2024-06-04 - Earliest claimed: 2021-12-08 - Assignees: Hyundai Motor Company; Kia Corporation - Applicants: Hyundai Motor Company; Kia Corporation - Inventors: Myoung Ok Lee; Jung Keun You; Mun Hwan Cho; Kaang Dok Yee - Web page: https://patentgrep.com/patent/US12002442B2 ## CPC - G10K 11/1754: {Speech masking} - G10L 17/26: Recognition of special voice characteristics, e.g. for use in lie detectors; Recognition of animal voices - G10L 25/18: the extracted parameters being spectral information of each sub-band - G10L 25/30: using neural networks - H04R 1/025: {Arrangements for fixing loudspeaker transducers, e.g. in a box, furniture} - G10L 17/02: Preprocessing operations, e.g. segment selection; Pattern representation or modelling, e.g. based on linear discriminant analysis [LDA] or principal components; Feature selection or extraction - G10L 17/18: Artificial neural networks; Connectionist approaches - H04R 3/00: Circuits for transducers (arrangements for producing a reverberation or echo sound G10K15/08; amplifiers H03F) - H04R 2499/13: Acoustic transducers and sound field adaptation in vehicles - G10K: SOUND-PRODUCING DEVICES; METHODS OR DEVICES FOR PROTECTING AGAINST, OR FOR DAMPING, NOISE OR OTHER ACOUSTIC WAVES IN GENERAL; ACOUSTICS NOT OTHERWISE PROVIDED FOR - G10L: SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING - H04R: LOUDSPEAKERS, MICROPHONES, GRAMOPHONE PICK-UPS OR LIKE ACOUSTIC ELECTROMECHANICAL TRANSDUCERS; ELECTRIC HEARING AIDS; PUBLIC ADDRESS SYSTEMS (producing sounds with frequency not determined by supply frequency G10K) ## Abstract The present disclosure provides a computer-implemented method for sound masking. The method includes identifying one class among a plurality of classes based on frequency characteristics of a voice of a first occupant in a vehicle, obtaining reference voice data corresponding to the identified class, generating a personalized masking sound for the first occupant by synthesizing the reference voice data with white noise, and controlling an output of the personalized masking sound. ## Claims ### Claim 1 A computer-implemented method for sound masking, the method comprising: identifying, by a controller, one class among a plurality of classes based on frequency characteristics of a voice of a first occupant in a vehicle; obtaining, by the controller, reference voice data corresponding to the identified class; generating, by the controller, a personalized masking sound for the first occupant by synthesizing the reference voice data with white noise; and controlling, by the controller, an output of the personalized masking sound. ### Claim 2 (depends on 1) The method of claim 1, wherein the plurality of classes are classified based on different frequency characteristics. ### Claim 3 (depends on 1) The method of claim 1, wherein the identifying of the one class among the plurality of classes comprises: acquiring, by the controller, a time-frequency representation of the voice; and identifying, by the controller, a class corresponding to the time-frequency representation among the plurality of classes as the one class. ### Claim 4 (depends on 3) The method of claim 3, wherein the identifying of the class corresponding to the time-frequency representation comprises: inputting, by the controller, the time-frequency representation into a classification model having been trained to classify a class of time-frequency representation of training voice data; and identifying, by the controller, the class corresponding to the time-frequency representation of the voice based on an output of the classification model. ### Claim 5 (depends on 4) The method of claim 4, wherein the classification model includes a convolutional neural network. ### Claim 6 (depends on 3) The method of claim 3, wherein the time-frequency representation is either a spectrogram or a mel-spectrogram. ### Claim 7 (depends on 1) The method of claim 1, wherein the controlling of the output of the personalized masking sound comprises: adjusting, by the controller, a magnitude of the personalized masking sound based on a magnitude of the voice. ### Claim 8 (depends on 1) The method of claim 1, wherein the controlling of the output of the personalized masking sound comprises: controlling, by the controller, at least one speaker such that the personalized masking sound is output to a second occupant. ### Claim 9 A sound masking device comprising: a microphone configured to receive a voice of a first occupant in a vehicle; at least one speaker disposed in the vehicle; and a controller configured to identify one class among a plurality of classes based on frequency characteristics of the voice, obtain reference voice data corresponding to the identified class, generate a personalized masking sound for the first occupant by synthesizing the reference voice data with white noise, and control the at least one speaker to output the personalized masking sound. ### Claim 10 (depends on 9) The sound masking device of claim 9, wherein the plurality of classes are classified based on different frequency characteristics. ### Claim 11 (depends on 9) The sound masking device of claim 9, wherein the controller is further configured to: acquire a time-frequency representation of the voice; and identify a class corresponding to the time-frequency representation among the plurality of classes as the one class. ### Claim 12 (depends on 11) The sound masking device of claim 11, wherein the controller is further configured to: input the time-frequency representation to a classification model having been trained to classify a class of time-frequency representation of training voice data; and identify the class corresponding to the time-frequency representation of the voice based on an output of the classification model. ### Claim 13 (depends on 12) The sound masking device of claim 12, wherein the classification model includes a convolutional neural network. ### Claim 14 (depends on 9) The sound masking device of claim 9, wherein the controller is configured to adjust a magnitude of the personalized masking sound based on a magnitude of the voice. ### Claim 15 (depends on 9) The sound masking device of claim 9, wherein the controller is configured to control the at least one speaker to output the personalized masking sound to a second occupant.