DETAILED ACTION
This action is in response to the communication filed 08/13/2026. Claims 11, 13-16 and 25-31 are pending.
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
Applicants’ arguments filed 08/13/2026 have been fully considered but they are not persuasive.
Regarding the rejections under 35 U.S.C 101
Applicant alleges that under step 2A prong 2 the claims as a whole recite additional elements that amount to significantly more. Applicant highlights the claims reflect and improvement to the functioning of a computer by providing in the claimed autoencoder an improved neural network. Further, Applicant cites the provisional application for support that such a network provides superior denoising performance and high efficiency.
Examiner disagrees.
First the claim amendment “the quadratic autoencoder providing a deep learning model with a reduced number of trainable parameters for the same or better denoising performance when compared to deep learning models based on linear neurons” can not be considered to reflect an improvement. The claim limitation merely recites the idea of a solution. Providing a network with an improved set of parameters compared to other models does not explain “how” the improvement is achieved. As noted in the MPEP “ if the specification explicitly sets forth an improvement but in a conclusory manner (i.e., a bare assertion of an improvement without the detail necessary to be apparent to a person of ordinary skill in the art), the examiner should not determine the claim improves technology”
As alleged by the Applicant, it is clear that the claimed autoencoder network may provide superior performance, however the improvement is the result of the recited abstract idea alone. The improvement is to the calculations performed by particular neurons, thus an improvement to the recited abstract idea. Further, as noted by the MPEP “the judicial exception alone cannot provide the improvement. The improvement can be provided by one or more additional elements”. It is clear that the improvement is not the result of any recited additional elements, but the computations performed by particular neurons.
In summary, while the claims recite the computations performed by the 2nd order neuron, the disclosure does not describe how this configuration affects an improvement. As required by the MPEP “a technical explanation of the asserted improvement is present in the specification” is required to assert the claims reflect an improvement. At most the disclosure provides a conclusory statement that such a network is superior.
Claim Rejections - 35 USC § 101
35 U.S.C. 101 reads as follows:
Whoever invents or discovers any new and useful process, machine, manufacture, or composition of matter, or any new and useful improvement thereof, may obtain a patent therefor, subject to the conditions and requirements of this title.
Claims 11, 13-16 and 25-31 are rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more.
Regarding Claim 11/26
Step 1 Analysis: Claim 11 is directed an image denoising system, which is directed to a machine, one of the statutory categories.
Step 1 Analysis: Claim 26 is directed A quadratic autoencoder trained on a set of images data to denoise an input image, which is directed to a machine, one of the statutory categories.
Step 2A Prong One Analysis: The claim recites a machine each of the following limitations:
determine a first dot product of an intermediate vector and an the input vector, the intermediate vector corresponding to a product of the input vector and the first weight vector or the input vector and the weight matrix… to determine a second dot product of the input vector and the second weight vector… determine the output of the second order neuron based, at least in part, on the first dot product and the second dot product
As drafted, is a machine that, under its broadest reasonable interpretation, recites mere instructions to implement an abstract idea on a computer. The above limitations in the context of this claim encompasses determining (mental processes). Determining an output of a dot product according to the mathematical equation is a mathematical concept that can be evaluated in a human mind. The additional limitations identified above only serve to describe the values which are used to perform the abstract idea.
As such the claim recites an abstract idea.
Step 2A Prong Two Analysis: The judicial exception is not integrated into a practical application. In particular, the claim only recited additional elements that are mere instructions to implement an abstract idea, or merely uses a computer as a tool to perform an abstract idea.
The additional elements (“a device comprising a processor circuitry, a memory circuitry and an artificial neural network (ANN) management circuitry; a quadratic autoencoder… wherein each second order neuron of the plurality of second order neurons is configured to implement a quadratic function of an input vector, wherein each second order neuron of the plurality of second order neurons comprises a first dot product circuitry comprising a first multiplier block and configured to… a second dot product circuitry comprising a second multiplier block and configured… and a nonlinear circuitry configured to”) amounts to mere instructions to implement an abstract ideas on a computer, or merely uses a computer as a tool to perform an abstract idea, as discussed in MPEP2106.05(f).
The additional elements (“incorporating second order neurons, the quadratic autoencoder providing a deep learning model with a reduced number of trainable parameters for the same or better denoising performance when compared to deep learning models based on linear neurons, wherein the quadratic autoencoder is trained on a training set of image data and validated on a validation set of image data to denoise an input image, the quadratic autoencoder”) amounts to the claim recites only the idea of a solution or outcome i.e., the claim fails to recite details of how a solution to a problem is accomplished. The recitation of claim limitations that attempt to cover any solution to an identified problem with no restriction on how the result is accomplished, as discussed in MPEP2106.05(f)(1))
In addition, the limitation (“comprising a convolutional layer and a deconvolutional layer, wherein each of the convolutional layer and the deconvolutional layer include a plurality of the second order neurons, each second order neuron of the plurality of second order neurons having an associated first weight vector and/or weight matrix, and a second weight vector, elements of the first weight vector, the weight matrix, and the second weight vector being determined by training on the training set of image data and validated on the validation set of image data… the nonlinear circuitry comprising a sigmoid block or a rectified linear unit block, the input vector, the intermediate vector, the first weight vector and the second weight vector each containing a number, n, elements, and the weight matrix having dimension n x n.)only generally links the use of the judicial exception to a particular technological environment, i.e. neural network processing. None of these limitations describe how any functions are performed only that labeled computer components comprise other components and are configured for or determined by training and/or denoising and/or validation without any detail with respect to how these functions are performed. (see MPEP 2106.05(h)).
Under step 2B: The recited additional elements when considered alone or in combination neither integrates the abstract idea into a practical application nor provides significantly more than the abstract idea itself.
Regarding Claim 13-16
Step 1 Analysis: The rejection of Claim 11 is incorporated, therefore Claim 13-16 is directed to a computer system, which is directed to a machine, one of the statutory categories.
Step 2A Prong One Analysis: Further, the claim recites a computer machine each of the following limitations:
determine a third dot product…multiply the second dot product and the third dot product…a summer circuitry configured to add the intermediate product and the first dot product…summer circuitry configured to add the first dot product and the second dot product to yield an intermediate output
As the rejection of claim 11 is incorporated, the claim recites an abstract idea.
Step 2A Prong Two Analysis:
The judicial exception is not integrated into a practical application. In addition to those additional elements already identified in the parent claim, the claim only recites additional elements that are mere instructions to implement an abstract idea, or merely uses a computer as a tool to perform an abstract idea. The additional element of a matrix, vectors, circuitry, multiplier blocks, summer blocks, and second order neurons amounts to mere instructions to implement an abstract idea on a computer, or merely uses a computer as a tool to perform an abstract idea, as discussed in MPEP2106.05(f). Further as noted previously in the rejection of claim 11 the additional element “the elements of the third weight vector being determined by training on the training set of image data and validation on the validation set of image data” only generally links the use of the judicial exception to a particular technological environment, i.e. neural network processing. The claims sets no limits on how the functions are performed. (see MPEP 2106.05(h)).
Step 2B: The recited additional elements when considered alone or in combination neither integrates the abstract idea into a practical application nor provides significantly more than the abstract idea itself.
Regarding Claim 25
Step 1 Analysis: The rejection of Claim 11 is incorporated, therefore Claim 25 is directed to a computer system, which is directed to a machine, one of the statutory categories.
Step 2A Prong One Analysis:
Further, the claim recites a computer machine each of the following limitations:
wherein an output of each second order neuron of the plurality of second order neurons corresponds to…
PNG
media_image1.png
247
645
media_image1.png
Greyscale
As the rejection of claim 11 is incorporated, the claim recites an abstract idea. Further, these limitations recite mathematical operations performed to arrive at an output for a neuron. As such, these limitations recite an additional abstract idea.
Step 2B:
The recited additional elements when considered alone or in combination neither integrates the abstract idea into a practical application nor provides significantly more than the abstract idea itself.
Regarding Claim 27-30
The dependent claims 27-30 are rejected for the same reasons as the dependent claims 13-16 in connection with claim 26.
Regarding Claim 31
The dependent claims 31 is rejected for the same reasons as the dependent claims 25 in connection with claim 26.
Allowable Subject Matter
The closest prior art of record is the newly cited Yang et al. “High Order Neural Networks with Reduced Numbers of Interconnection Weights” which describes the generalized equation for a higher order neuron
PNG
media_image2.png
48
332
media_image2.png
Greyscale
and specifically the equation for a parabolic neuron
PNG
media_image3.png
57
439
media_image3.png
Greyscale
.
The cited art either alone or in combination do not teach all of the limitations of claims 11 and 26.
Conclusion
Prior art
Lu et al. “An efficient multilayer quadratic perceptron for pattern classification and function approximation” teaches a sigma pi style neural unit for implementing a neural unit for non-linear function approximation.
Ganesh et al “Pattern Classification using Quadratic Neuron: An Experimental Study” teaches a quadratic neuron model for classification of concentric separable classes.
Su et al. “A Neural-Network-Based Approach to Detecting Hyperellipsoidal Shells” disclosed neural networks that implement quadratic junctions in order to learn hyper spherical patterns.
Yang et al. “High Order Neural Networks with Reduced Numbers of Interconnection Weights” discloses a generalized mathematical form for higher order neurons in neural networks. It also specifically provides an example equation for a parabolic neuron:
PNG
media_image3.png
57
439
media_image3.png
Greyscale
THIS ACTION IS MADE FINAL. Applicant is reminded of the extension of time policy as set forth in 37 CFR 1.136(a).
A shortened statutory period for reply to this final action is set to expire THREE MONTHS from the mailing date of this action. In the event a first reply is filed within TWO MONTHS of the mailing date of this final action and the advisory action is not mailed until after the end of the THREE-MONTH shortened statutory period, then the shortened statutory period will expire on the date the advisory action is mailed, and any extension fee pursuant to 37 CFR 1.136(a) will be calculated from the mailing date of the advisory action. In no event, however, will the statutory period for reply expire later than SIX MONTHS from the mailing date of this final action.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to JOHNATHAN R GERMICK whose telephone number is (571)272-8363. The examiner can normally be reached on Monday-Friday 7:30 am – 4:00 pm (EST).
If attempts to reach the examiner by telephone are unsuccessful, the examiner’s supervisor, Kakali Chaki, can be reached at telephone number 5712723719. The fax phone number for the organization where this application or proceeding is assigned is 571-273-8300.
Information regarding the status of an application may be obtained from the Patent Application Information Retrieval (PAIR) system. Status information for published applications may be obtained from either Private PAIR or Public PAIR. Status information for unpublished applications is available through Private PAIR only. For more information about the PAIR system, see http://portal.uspto.gov/external/portal. Should you have questions about access to the Private PAIR system, contact the Electronic Business Center (EBC) at 866-217-9197 (toll-free).
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.
/J.R.G./ Examiner, Art Unit 2122 /KAKALI CHAKI/Supervisory Patent Examiner, Art Unit 2122
Prosecution Insights