DETAILED ACTION
Notice of Pre-AIA or AIA Status
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
Response to Arguments
Applicant's arguments filed 15 April 2026 have been fully considered but they are not persuasive.
Applicant argues that Fitz does not teach a DNN processing loop configured to determine a denoiser mask associated with a noise prediction for a version of the audio signal sample that is received by the DNN processing loop prior to the delay being added via the time-frequency domain transformation pipeline. However, the binary mask described by Fitz is indeed associated with a noise prediction (see [0016], [0019], [0021]: the binary mask is based on an estimate of noise), and per [0034] and Figure 3, the binary mask is applied to a version of the signal without delay added, as the delay is only in the unenhanced signal branch. Applicant also argues that Fitz does not disclose a DNN processing loop, which is true, however this a piecemeal analysis of the references. Fitz is invoked in the rejection to read on the implementation of the delay unit in relation to the other structural components of the invention; the details of the other components of the invention, such as the use of DNN processing to acquire the mask, are taught by references such as Crow elsewhere in the rejection. One cannot show nonobviousness by attacking references individually where the rejections are based on combinations of references. See In re Keller, 642 F.2d 413, 208 USPQ 871 (CCPA 1981); In re Merck & Co., 800 F.2d 1091, 231 USPQ 375 (Fed. Cir. 1986).
Claim Rejections - 35 USC § 103
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
Claim(s) 1-8, 10-15, 17-21 is/are rejected under 35 U.S.C. 103 as being unpatentable over Crow et al. (US 20190082276 A1) in view of Zhang et al. (US 20190080710 A1) in view of Fitz (US 20150092967 A1).
Regarding claim 1, Crow et al. teaches a digital signal processing (DSP) apparatus configured to reduce noise from an audio signal sample associated with at least one microphone ([0016]-[0017], FIG 2 audio sensors 212), the DSP apparatus comprising one or more processors and one or more memories instructions that (FIG 2, processors and storage)
input the audio signal sample to a both a time-frequency domain transformation pipeline and a deep neural network (DNN) processing loop, wherein the time-frequency domain transformation pipeline forms part of a digital signal processing process and configured to execute one or more transformation pipeline steps of the time-frequency domain transformation pipeline during a transformation time period, ([0026]-[0028]: collect audio data from audio sensors, provided to the system for processing including time/frequency domain conversion [0035]-[0040], applying neural network models, time-bounded filter/frequency coefficients, [0052]-[0053] frequency masks)
wherein the DNN processing loop is configured to (i) determine a denoiser mask associated with a noise prediction for the audio signal sample ([0035]-[0040], applying neural network models, time-bounded filter/frequency coefficients, [0052]-[0053] frequency masks)
based on the denoiser mask being determined prior to an expiration of the transformation time period, output a DNN denoised version of the audio signal sample via the time-frequency domain transformation pipeline by applying the denoiser mask to the audio signal sample; ([0039], [0046], time-bounded filter; filters are applied respective to the time when the audio is received and to a “time period of validity”, thus if within the time period of validity, the filter will be applied)
Crow et al does not disclose wherein the DNN processing loop is configured to (ii) utilize machine learning to determine the denoiser mask at least approximately in parallel to the one or more transformation pipeline steps performed by the time-frequency domain transformation pipeline;
However, Zhang et al. does disclose wherein the DNN processing loop is configured to (ii) utilize machine learning to determine the denoiser mask at least approximately in parallel to the one or more transformation pipeline steps performed by the time-frequency domain transformation pipeline; (Figure 4, [0007], claims 1-4: Using feature learning to predict and identify noise components and output a matrix to serve as a denoiser mask in between time frequency transformation steps; processing unit 204 inherently enables multithreaded implementation; Figure 3: starting from block 322, two parallel paths are shown- one path uses the audio sample to determine the denoiser mask; while the other path provides the time/frequency transformed sample, STFT is an appropriate step for both pipelines)
based the denoiser mask not being determined prior to the expiration of the transformation time period, output a default version of the audio signal sample via the time- frequency domain transformation pipeline without applying the denoiser mask to the audio signal sample. ([0046]: time-bounded filters have a limited time period of validity to which they can be applied to the signal; [0072]: use a default earpiece filter when there isn’t a new time-valid filter available)
It would have been obvious to one with ordinary skill in the art before the effective filing date incorporate the DNN processing steps of Zhang et al. to produce a denoiser mask because by isolating noise from speech and creating a mask using deep learning techniques, the intelligibility of speech in audio can be enhanced (See Zhang et al. [0007])
Crow and Zhang do not teach wherein the time-frequency domain transformation pipeline forms part of digital signal processing process and is configured to add a delay to the audio signal sample,
wherein the DNN processing loop is configured to determine a denoiser mask associated with a noise prediction for a version of the audio signal sample that is received by the DNN processing loop prior to the delay being added via the time-frequency domain transformation pipeline.
However, Fitz does teach wherein the time-frequency domain transformation pipeline forms part of digital signal processing process and is configured to add a delay to the audio signal sample. (Figure 3, [0034]: in a parallel unenhanced signal branch, include a delay unit)
wherein the DNN processing loop is configured to determine a denoiser mask associated with a noise prediction for a version of the audio signal sample that is received by the DNN processing loop prior to the delay being added via the time-frequency domain transformation pipeline. (Figure 3, [0034]: a denoiser mask is determined in the enhancement signal branch, which does not have a delay unit)
It would have been obvious to one with ordinary skill in the art before the effective filing date of the claimed invention to include a delay unit in a parallel branch as taught by Fitz in the teachings of Crow and Zhang because the delay unit compensates for processing latency introduced in the parallel enhancement branch (see Fitz [0034]).
Regarding claim 2, elements of parent claim 1 are disclosed as written above. Crow et al. further teaches the DSP apparatus wherein the instructions are further operable to cause the DSP apparatus to:
based on the denoiser mask not being determined prior to expiration of the transformation period, apply a default denoiser mask associated with a default noise prediction to the audio signal sample associated with the time-frequency domain transformation pipeline. ([0072] default earpiece filter)
Regarding claim 3, elements of parent claim 1 are disclosed as written above. Crow et al. further teaches the DSP apparatus wherein the instructions are further operable to cause the DSP apparatus to:
based on the denoiser mask not being determined prior to expiration of the transformation period, apply a prior denoiser mask associated with a prior noise prediction to the audio signal sample associated with the time-frequency domain transformation pipeline. ([0061] using recently used filters)
Regarding claim 4, elements of parent claim 3 are disclosed as written above. Crow et al. further teaches the DSP apparatus wherein the instructions are further operable to cause the DSP apparatus to: modify the prior denoiser mask in response to applying the prior denoiser mask to the audio signal sample associated with the time- frequency domain transformation pipeline. ([0061] create a best estimate filter by averaging previously used filters)
Regarding claim 5, elements of parent claim 1 are disclosed as written above. Crow et al. further teaches the DSP apparatus wherein the instructions are further operable to cause the DSP apparatus to:
based on the denoiser mask not being determined prior to expiration of the transformation period, apply a passthrough denoiser mask configured without denoising to the audio signal sample associated with the time-frequency domain transformation pipeline. ([0072] store any number of fallback filter types)
Regarding claim 6, elements of parent claim 1 are disclosed as written above. Crow et al. further teaches the DSP apparatus wherein the instructions are further operable to cause the DSP apparatus to:
based on the denoiser mask not being determined prior to expiration of the transformation period, apply a band-pass shape denoiser mask to the audio signal sample associated with the time-frequency domain transformation pipeline. ([0072] fallback filters include a band-pass filter)
Regarding claim 7, elements of parent claim 1 are disclosed as written above. Crow et al. further teaches the DSP apparatus wherein the instructions are further operable to cause the DSP apparatus to:
based on the denoiser mask not being determined prior to expiration of the transformation period, apply a low-pass shape denoiser mask to the audio signal sample associated with the time-frequency domain transformation pipeline. ([0072] fallback filters include a low-pass filter)
Regarding claim 8, elements of parent claim 1 are disclosed as written above. Crow et al. further teaches the DSP apparatus wherein the instructions are further operable to cause the DSP apparatus to: receive user denoiser control parameters; ([0011], [0076]: user preference data)
apply the user denoiser control parameters to the denoiser mask to generate a user-modified denoiser mask; ([0039], [0072] user feedback, preference data)
based on the user-modified denoiser mask being determined prior to expiration of the transformation period, apply the user-modified denoiser mask to the frequency domain version of the audio signal sample associated with the time-frequency domain transformation pipeline to generate a user-modified denoised audio signal sample. ([0039] time-bounded filter)
Regarding claim 10, elements of parent claim 1 are disclosed as written above. Crow et al. further teaches the DSP apparatus wherein the denoiser mask is applied to a first frequency domain audio signal sample, and wherein the instructions are further operable to cause the DSP apparatus to: transform the audio signal sample into the first frequency domain audio signal sample via the time-frequency domain transformation pipeline; ([0025]-[0028] frequency domain conversion)
and transform the audio signal sample into a second frequency domain audio signal sample via the DNN processing loop. ([0025]-[0028], [0035] applying neural network model)
provide the second frequency domain audio signal sample to a DNN model that is configured to determine the denoiser mask; ([0035]-[0040], applying neural network model)
based on the denoiser mask being determined prior to expiration of the transformation period, apply the denoiser mask to the first frequency domain audio signal sample. ([0039] time-bound filter)
Regarding claim 11, elements of parent claim 1 are disclosed as written above. Crow et al. further teaches the DSP apparatus wherein the instructions are further operable to cause the DSP apparatus to: modify frequency of one or more portions of the audio signal sample to generate a modified audio signal sample; ([0028]-[0030], frequency pre-processing)
transform the modified audio signal sample into a frequency domain audio signal sample; modify frequency of one or more portions of the frequency domain audio signal sample to generate a modified frequency domain audio signal sample; ([0025]-[0028] frequency domain conversion)
and determine the denoiser mask associated with the noise prediction based on the modified frequency domain audio signal sample. ([0035]-[0040], applying neural network models, time-bounded filter/frequency coefficients, [0052]-[0053] frequency masks)
Regarding claim 12 elements of parent claim 1 are disclosed as written above. Crow et al. further teaches the DSP apparatus wherein the instructions are further operable to cause the DSP apparatus to: provide the modified frequency domain audio signal sample to a DNN model that is configured to determine the denoiser mask; ([0035]-[0040], applying neural network models, time-bounded filter/frequency coefficients, [0052]-[0053] frequency masks)
based on the denoiser mask being determined prior to expiration of the transformation period, apply the denoiser mask to the audio signal sample associated with the time-frequency domain transformation pipeline. ([0039], time-bounded filter)
Regarding claim 13, elements of parent claim 1 are disclosed as written above. Crow et al. further teaches the DSP apparatus wherein the instructions are further operable to cause the DSP apparatus to: perform spatial filtering of the denoiser mask to generate an optimized denoiser mask; ([0058], [0061]: spatial filtering)
based on the optimized denoiser mask being determined prior to expiration of the transformation period, apply the optimized denoiser mask to the audio signal sample associated with the time-frequency domain transformation pipeline to generate an optimized denoised audio signal sample associated with the at least one microphone. ([0039], time-bounded filter)
Regarding claim 14, elements of parent claim 1 are disclosed as written above. Crow et al. further teaches the DSP apparatus wherein the instructions are further operable to cause the DSP apparatus to: apply the denoiser mask associated with the noise prediction to the audio signal sample associated with the time-frequency domain transformation pipeline based on a user bypass input parameter associated with the time-frequency domain transformation pipeline satisfying a defined bypass criterion. ([0031], [0076], [0079]: receive user input for feedback or user-defined parameters; raw audio mode)
Regarding claim 15, elements of parent claim 1 are disclosed as written above. Crow et al. further teaches the DSP apparatus wherein the instructions are further operable to cause the DSP apparatus to: provide the audio signal sample to a DNN model of the DNN processing loop that is configured to predict whether the audio signal sample includes one or more signals of interest and to configure the denoiser mask based on the one or more signals of interest; ([0035]-[0040], applying neural network models, detecting trigger conditions)
based on the denoiser mask being determined prior to expiration of the transformation period, scale active noise cancellation associated with the audio signal sample based on the denoiser mask. ([0039], time-bounded filter)
Regarding claim 17, it is analogous to claim 1 and is rejected in a similar fashion.
Regarding claim 18, elements of parent claim 17 are disclosed as written above. Crow et al. further teaches the computer-implemented method:
wherein outputting the default version of the audio signal sample via the time-frequency domain transformation comprises applying a default denoiser mask associated with a default noise prediction to the audio signal sample associated with the time-frequency domain transformation pipeline. ([0072] storing a default earpiece filter or default filter parameters)
Regarding claim 19, elements of parent claim 17 are disclosed as written above. Crow et al. further teaches the computer-implemented method further comprising:
wherein outputting the default version of the audio signal sample via the time-frequency domain transformation pipeline comprises applying a predicted denoiser mask to the audio signal sample associated with the time-frequency domain transformation pipeline. ([0061] infer possible filters in case of connection problems)
Regarding claim 20, elements of parent claim 17 are disclosed as written above. Crow et al. further teaches the computer-implemented method further comprising:
wherein outputting the default version of the audio signal sample via the time-frequency domain transformation pipeline comprises applying a prior denoiser mask configured without denoising to the audio signal sample associated with the time- frequency domain transformation pipeline. ([0061] using recently used filters)
Regarding claim 21, it is analogous to claim 1 and is rejected in a similar fashion.
Regarding claim 22, elements of parent claim 1 are disclosed as written above. Crow et al. further teaches the computer-implemented method further comprising:
generate the default version of the audio signal sample based on a default denoiser mask associated with a default noise prediction for the time-frequency domain transformation pipeline, a prior denoiser mask associated with a prior noise prediction provided by the DNN processing loop, a passthrough mask for the time-frequency domain transformation pipeline that is configured without denoising, a band-pass shape denoiser mask associated with band-pass filtering, a low-pass shape denoiser mask associated with low-pass filtering, or default DSP processing associated with the time-frequency domain transformation pipeline. ([0072]: storing default filter parameters, cached filter parameters, high/low/band pass, etc.)
Claim(s) 9 is/are rejected under 35 U.S.C. 103 as being unpatentable over Crow et al in view of Zhang et al. as applied to claim 1 above, and further in view of Vilermo et al. (US 20190088267 A1).
Regarding claim 9, elements of parent claim 1 are disclosed as written above.. Crow et al. does not disclose the DSP apparatus wherein the instructions are further operable to cause the DSP apparatus to: generate a dynamic noise reduction interface object that is configured to cause a client device to render a dynamic noise reduction interface to visually indicate a degree of noise reduction provided by the denoiser mask;
and output the dynamic noise reduction interface object to the client device.
Villermo et al. does disclose the DSP apparatus wherein the instructions are further operable to cause the DSP apparatus to: generate a dynamic noise reduction interface object that is configured to cause a client device to render a dynamic noise reduction interface to visually indicate a degree of noise reduction provided by the denoiser mask; ([0113], [0116]: render a menu that displays and allows the user to select a level of noise reduction)
and output the dynamic noise reduction interface object to the client device. ([0112]-[0113]: displaying noise reduction menu)
While Crow et al. discloses a client device that can take user input to interact with the noise reduction system, it does not specify an interface object to indicate degrees of noise reduction. Villermo et al. does teach this concept, and it would have been obvious to one with ordinary skill in the art before the effective filing date to include a noise reduction interface taught by Villermo et al. because it enables the user to control the noise reduction systems supplied by the invention (Villermo et al. [0110]).
Claim(s) 16 is/are rejected under 35 U.S.C. 103 as being unpatentable over Crow et al. in view of Zhang et al. as applied to claim 1 above, and further in view of Baby et al. ("Machines hear better when they have ears").
Regarding claim 16, elements of parent claim 1 are disclosed as written above. Crow et al. does not disclose the DSP apparatus wherein the audio signal sample is an otoacoustic emissions signal sample, and wherein the instructions are further operable to cause the DSP apparatus to: provide the otoacoustic emissions signal sample to a DNN model of the DNN processing loop that is configured to predict whether the otoacoustic emissions signal sample includes one or more signals of interest and to configure the denoiser mask based on the one or more signals of interest.
Baby et al. does disclose the DSP apparatus wherein the audio signal sample is an otoacoustic emissions signal sample, and wherein the instructions are further operable to cause the DSP apparatus to: provide the otoacoustic emissions signal sample to a DNN model of the DNN processing loop that is configured to predict whether the otoacoustic emissions signal sample includes one or more signals of interest and to configure the denoiser mask based on the one or more signals of interest. (pages 4-5, nonlinear transmission line model: simulated otoacoustic emissions signals are input to a noise reduction DNN)
While Crow et al. includes inputting audio signals to a DNN as part of a noise reduction process, it does not mention inputting otoacoustic emission signals. Baby et al. does teach this concept, and it would have been obvious to one with ordinary skill in the art before the effective filing date to include otoacoustic emissions signals as an input because including biophysical features can improve the effectiveness of speech enhancement and noise reduction systems (Baby et al. page 1, 5).
Conclusion
Any inquiry concerning this communication or earlier communications from the examiner should be directed to ALVIN ISKENDER whose telephone number is (703)756-4565. The examiner can normally be reached M-F.
Examiner interviews are available via telephone, in-person, and video conferencing using a USPTO supplied web-based collaboration tool. To schedule an interview, applicant is encouraged to use the USPTO Automated Interview Request (AIR) at http://www.uspto.gov/interviewpractice.
If attempts to reach the examiner by telephone are unsuccessful, the examiner’s supervisor, HAI PHAN can be reached on (571) 272-6338. The fax phone number for the organization where this application or proceeding is assigned is 571-273-8300.
Information regarding the status of published or unpublished applications may be obtained from Patent Center. Unpublished application information in Patent Center is available to registered users. To file and manage patent submissions in Patent Center, visit: https://patentcenter.uspto.gov. Visit https://www.uspto.gov/patents/apply/patent-center for more information about Patent Center and https://www.uspto.gov/patents/docx for information about filing in DOCX format. For additional questions, contact the Electronic Business Center (EBC) at 866-217-9197 (toll-free). If you would like assistance from a USPTO Customer Service Representative, call 800-786-9199 (IN USA OR CANADA) or 571-272-1000.
/ALVIN ISKENDER/Examiner, Art Unit 2654
/HAI PHAN/Supervisory Patent Examiner, Art Unit 2654
Prosecution Insights