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 .
Continued Examination Under 37 CFR 1.114
A request for continued examination under 37 CFR 1.114, including the fee set forth in 37 CFR 1.17(e), was filed in this application after final rejection. Since this application is eligible for continued examination under 37 CFR 1.114, and the fee set forth in 37 CFR 1.17(e) has been timely paid, the finality of the previous Office action has been withdrawn pursuant to 37 CFR 1.114. Applicant's submission filed on 01/20/2026 has been entered.
Response to Arguments
Applicant’s argument filed 01/20/2026 have been fully considered but they are not persuasive. The amendments have overcome 112 rejection and thus, 112 rejections have been withdrawn.
Applicant’s Argument: On page 7-8 of Applicant’s response to rejections under 35 U.S.C. 101, applicant states “The claimed invention improves the technical field of data encoding and physical modeling by allowing a system to predict a physical phenomenon for different data modalities based on a unimodal input. As such, the cross-modal inference employed by the claimed invention allows simulation of high-fidelity, low-throughput measurements from low-fidelity, high-throughput measurements. (See Specification, [0048].) This ability allows the system to accurately predict physical phenomena when limited data (and data modes) are available.”
Examiner’s Response: Applicant’s argument is not persuasive. An important consideration in determining whether a claim improves technology is the extent to which the claim covers a particular solution to a problem or a particular way to achieve a desired outcome, as opposed to merely claiming the idea of a solution or outcome (see MPEP 2106.05(a)). The amended claims do not provide sufficient details to describe any technological improvement. If the specifications explicitly set forth an improvement but in a conclusory manner (see MPEP 2106.04(d)(1): 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.
Applicant states that the claimed improvement of the cross-modal inference allows the system to accurately predict physical phenomena when limited data are available. The claims as recited do not provide any details on how the cross-modal inference is performed. Additionally, the encoding and decoding steps are recited in a broad manner without describing any particular way to achieved the desired outcome.
Applicant’s Argument: On page 8-10 of Applicant’s response to rejections under 35 U.S.C. 103, applicant states “However, Takeishi does not disclose or suggest a Gaussian mixture of latent representations using parallel decoders, even when physics knowledge is involved. A close reading of Takeishi reveals that it teaches a single variational autoencoder with a single decoder architecture. In Takeishi, the latent space is split, not mixed. (See Takeishi, p. 2.) The latent space is divided into two distinct parts: physics-grounded latent variable (ZP) tied to parameters of the physics model, and auxiliary latent variables (ZA) handled by neural networks to capture effects not explained by the physics. Both ZP and ZA are assumed to follow Gaussian priors, but they are single Gaussians, not a Gaussian mixture. Furthermore, Takeishi discloses one decoder pipeline that combines the physics model and the neural network correction model. This decoder pipeline does not run as parallel decoders.”
Examiner’s Response: Applicant’s argument is not persuasive. In response to applicant's arguments against the references individually, 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).
Takeishi does not explicitly disclose the combinations of the individual Gaussian priors. However, Yakut (par. 114) discloses the combination of individual modalities into an approximate joint posterior and teaches the claim limitation of a Gaussian mixture. Yakut (par. 122 and Fig. 2B) also teaches a number of decoders that are in parallel connections with one another. Takeishi (pg. 1-2, Section 1, par. 4) discloses a decoder that comprises of physics-based models and trainable neural networks to reconstruct data. It would be obvious to one of the ordinary skill in the art before the effective filing date of the claimed invention to incorporate the teaching of a physics-constraint decoder from Takeishi into the teaching of Yakut because a physics-constraint decoder provides meaningful physics-based inductive bias to improve the performance of the generative model to extrapolate beyond the training dataset (Takeishi, abstract).
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 1-21 are rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more.
Regarding Claim 1:
Subject Matter Eligibility Analysis Step 1:
Claim 1 recites “A computer-implemented method of data encoding, the method comprising” and is thus a process, one of the four statutory categories of patentable subject matter.
Subject Matter Eligibility Analysis Step 2A Prong 1:
“encoding each of the different modalities of data into an individual latent representation” (a mental process that can be performed in the human mind with the aid of pen and paper, i.e. judgement)
“combining the individual latent representations into a single Gaussian mixture distribution in a shared latent space” (a mathematical relationship, i.e. conversion of representations into a distribution; See Spec. par. 25)
“decoding the Gaussian mixture ” (a mental process that can be performed in the human mind with the aid of pen and paper, i.e. judgement)
“predicting, from the unimodal dataset, a value of the physical phenomenon for the different modalities of data according to cross-modal inference learning from encoding and decoding of the multimodal dataset” (a mental process that can be performed in the human mind with the aid of pen and paper, i.e. judgement)
Claim 1 therefore recites an abstract idea.
Subject Matter Eligibility Analysis Step 2A Prong 2:
“using a number of processors to perform the steps of” (mere instructions to apply the exception using a generic computer component - see MPEP 2106.05(f))
"receiving a multimodal dataset comprising number of different modalities of data related to a physical phenomenon common to the different modalities of data” (This step is directed to data gathering, which is understood to be insignificant extra solution activity - see MPEP 2106.05(g))
“decoding knowledge and provide physics-based inductive biases, wherein the decoders and physics simulators respectively reconstruct the multimodal dataset” (mere instructions to apply the exception using a generic computer component - see MPEP 2106.05(f))
"receiving a unimodal dataset comprising a single modality of data related to the physical phenomenon” (This step is directed to data gathering, which is understood to be insignificant extra solution activity - see MPEP 2106.05(g))
The additional elements as disclosed above alone or in combination do not integrate the judicial exception into practical application as they are mere insignificant extra solution activity in combination of generic computer functions being implemented with generic computer elements in a high level of generality to perform the disclosed abstract idea above. Therefore, Claim 1 is directed to the abstract idea.
Subject Matter Eligibility Analysis Step 2B:
“using a number of processors to perform the steps of” (mere instructions to apply the exception using a generic computer component - see MPEP 2106.05(f))
"receiving a multimodal dataset comprising number of different modalities of data related to a physical phenomenon common to the different modalities of data” (This step is directed to transmitting or receiving information, which is understood to be insignificant extra solution activity and well understood, routine and conventional activity of transmitting and receiving data as identified by the court - see MPEP 2106.05(d))
“decoding physics simulators, wherein the physics simulators encode prior physics knowledge and provide physics-based inductive biases, wherein the decoders and physics simulators respectively reconstruct the multimodal dataset” (mere instructions to apply the exception using a generic computer component - see MPEP 2106.05(f))
"receiving a unimodal dataset comprising a single modality of data related to the physical phenomenon” (This step is directed to transmitting or receiving information, which is understood to be insignificant extra solution activity and well understood, routine and conventional activity of transmitting and receiving data as identified by the court - see MPEP 2106.05(d))
The additional elements as disclosed above alone or in combination do not recite significantly more than the abstract idea itself as they are mere insignificant extra solution activity in combination of generic computer functions being implemented with generic computer elements in a high level of generality to perform the disclosed abstract idea above. Therefore, Claim 1 is subject-matter ineligible.
Regarding Claim 8:
The claim recites a system (“A system for data autoencoding, the system comprising”) that performs the method as described in claim 1. Therefore, claim 8 is rejected for the same reasons as disclosed for claim 1. The limitations for additional elements of claim 8 are analyzed below.
Subject Matter Eligibility Analysis Step 2A Prong 1:
Please see Step 2A Prong 1 analysis of claim 1
Subject Matter Eligibility Analysis Step 2A Prong 2 & 2B:
“a storage device configured to store program instructions” (mere instructions to apply the exception using a generic computer component - see MPEP 2106.05(f))
“one or more processors operably connected to the storage device and configured to execute the program instructions to cause the system to” (mere instructions to apply the exception using a generic computer component - see MPEP 2106.05(f))
Regarding Claim 15:
Subject Matter Eligibility Analysis Step 1:
The claim recites an article of manufacture (“A computer program product for data autoencoding, the computer program product comprising”) that performs the method as described in claim 1. Therefore, claim 15 is rejected for the same reasons as disclosed for claim 1. The limitations for additional elements of claim 15 are analyzed below.
Subject Matter Eligibility Analysis Step 2A Prong 1:
Please see Step 2A Prong 1 analysis of claim 1
Subject Matter Eligibility Analysis Step 2A Prong 2 & 2B:
“a computer-readable storage medium having program instructions embodied thereon to perform the steps of” (mere instructions to apply the exception using a generic computer component - see MPEP 2106.05(f)).
Regarding Claims 2, 9, and 16:
Subject Matter Eligibility Analysis Step 2A Prong 1: None
Subject Matter Eligibility Analysis Step 2A Prong 2 & 2B:
“wherein the Gaussian mixture comprises a combination of clusters of sub-populations of the data, wherein the clusters represent all the modalities of data” (merely specifies a particular technological environment in which the abstract idea is to take place, ie. a field of use, and thus does not integrate the abstract idea into a practical application nor cannot provide significantly more than the abstract idea itself - see MPEP 2106.05(h))
Regarding Claims 3, 10, and 17:
Subject Matter Eligibility Analysis Step 2A Prong 1:
“wherein the clusters encode cross-modal shared information” (a mathematical calculation, i.e. conversion of data into coded representation)
Subject Matter Eligibility Analysis Step 2A Prong 2 & 2B: None
Regarding Claims 4, 11, and 18:
Subject Matter Eligibility Analysis Step 2A Prong 1: None
Subject Matter Eligibility Analysis Step 2A Prong 2 & 2B:
“wherein different clusters have different parameters for a same physics model” (merely specifies a particular technological environment in which the abstract idea is to take place, ie. a field of use, and thus does not integrate the abstract idea into a practical application nor cannot provide significantly more than the abstract idea itself - see MPEP 2106.05(h))
Regarding Claims 5, 12, and 19:
Subject Matter Eligibility Analysis Step 2A Prong 1: None
Subject Matter Eligibility Analysis Step 2A Prong 2 & 2B:
“wherein each modality of data is represented by a separate physics simulator among the physics simulators” (merely specifies a particular technological environment in which the abstract idea is to take place, ie. a field of use, and thus does not integrate the abstract idea into a practical application nor cannot provide significantly more than the abstract idea itself - see MPEP 2106.05(h))
Regarding Claims 6, 13, and 20:
Subject Matter Eligibility Analysis Step 2A Prong 1: None
Subject Matter Eligibility Analysis Step 2A Prong 2 & 2B:
“wherein the encoding and decoding comprise unsupervised learning” (mere instructions to apply the exception using a generic computer component - see MPEP 2106.05(f); unsupervised learning is a generic process performed by a generic ML model)
Regarding Claims 7, 14, and 21:
Subject Matter Eligibility Analysis Step 2A Prong 1: None
Subject Matter Eligibility Analysis Step 2A Prong 2 & 2B:
“wherein the Gaussian mixture is generated by a Product of Experts model” (mere instructions to apply the exception using a generic computer component - see MPEP 2106.05(f))
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.
Claims 1-5, 7-12, 14-19, and 21 are rejected under 35 U.S.C. 103 as being unpatentable over Yakut (US20230045548A1), in view of Takeishi, “Physics-Integrated Variational Autoencoders for Robust and Interpretable Generative Modeling”.
Regarding claim 1, Yakut teaches:
“A computer-implemented method of data encoding, the method comprising: using a number of processors to perform the steps of” (abstract, [0098-0099; 0142-0143, Figure 2A], A generative model is trained on data comprising of multiple modalities using a device that contains multiple processing units. Figure 2A shows the MVAE model and the system may be used to analyze an example of catalyst deactivation.)
“receiving a multimodal dataset comprising number of different modalities of data related to a physical phenomenon common to the different modalities of data” ([0010-0013, 0142-0148], Dataset related to catalyst deactivation (physical phenomenon) is received by the system and the data is divided into multiple modalities. The input data may consist of one or more condition parameters and one or more KPIs that may be represented as separate modality.)
“encoding each of the different modalities of data into an individual latent representation” ([0111, Figure 2A], Multimodal data contain N different modalities are input into individual encoders to generate a latent representation of the data.)
“combining the individual latent representations into a single Gaussian mixture distribution in a shared latent space” ([0114], Individual modalities are all Gaussian distributions and product of experts combines all individual modalities into a joint posterior.)
“decoding the Gaussian mixture with a number of parallel decoders and ” ([0122; 0135; 0137, Figure 2B & 3], The latent state from the joint posterior is fed into individual decoder networks to reconstruct the individual modalities. The plurality of decoders is shown in parallel connections. Figure 3 shows the MVAE system that consists of encoders, decoders, and product of experts. The system is configured to handle any number of missing modalities based on the given modalities and in some embodiments, the system can generate all the missing modalities from a single modality. The MVAE model may only have condition parameters provided as input data and the generative model will produce KPIs that are related to the given condition parameters.)
“receiving a unimodal dataset comprising a single modality of data related to the physical phenomenon” ([0012, 0041, 0137-0139, Figure 2B & 3], The system is able to function based on any provided combinations of modalities. In one embodiment, condition parameters of a single modality may be received as input and the system generates a prediction for one or more KPIs.)
“predicting, from the unimodal dataset, a value of the physical phenomenon for the different modalities of data according to cross-modal inference learning from encoding and decoding of the multimodal dataset” ([0012, 0018, 0041, 0137-0139, 0182-0184, Figure 2B & 3], The model may use input values for fewer modalities of data than the number of modalities used to train the generative model. In one embodiment, condition parameters that are aggregated into one modality may be received as input and the system generates a prediction for one or more KPIs. The model is trained on a combination of modalities to learn the relationships between modalities and the trained model can predict values of missing modalities.)
Yakut does not explicitly disclose an implementation of “a number of parallel decoders and physics simulators, wherein the physics simulators encode prior physics knowledge and provide physics-based inductive biases, wherein the decoders and physics simulators respectively reconstruct the multimodal dataset”. However, Takeishi discloses in the same field of endeavor:
“decoding the Gaussian physics knowledge and provide physics-based inductive biases, wherein the” ([pg. 1-2, Section 1, par. 4, Section 2.1; pg. 5, Section 3.2, par. 1-2; pg. 10, Section 5.4, par. 1-3], Physics models are integrated into variational autoencoders, where the decoder comprises physics-based models and trainable neural network. In the experiments, the movement of human subjects are modeled using a Hamiltonian neural network, which learns conversation laws from the locomotion data. The physics model is used as a source of information for data augmentation. The physics-integrated VAE reconstructs a test sample of the gait data. Under the broadest reasonable interpretation, the physics simulators reconstruct the dataset and the physics simulators can be considered as a decoder that has physics information added to the dataset. One or more of the parallel decoders from Yakut can be replaced with a decoder that comprises of physics-based models to provide physics-based inductive biases.)
It would be obvious to one of the ordinary skill in the art before the effective filing date of the claimed invention to incorporate the teaching of “a number of parallel decoders and physics simulators, wherein the physics simulators encode prior physics knowledge and provide physics-based inductive biases, wherein the decoders and physics simulators respectively reconstruct the multimodal dataset” from Takeishi into the teaching of Yakut by replacing one of the decoder components in Yakut with the physics-constraint decoder. Doing so can improve models with improved interpretability and abilities to extrapolate by integrating physics models into ML model (Takeishi, abstract).
Regarding claim 8:
Claim 8 recites a system (“A system for data autoencoding, the system comprising”) that performs the same process as described in Claim 1. Therefore claim 8 is rejected under the same reasons mention for claim 1. The additional elements of claim 8 is addressed below with Yakut:
“a storage device configured to store program instructions” ([0204-0205], Instructions are stored on a computer readable storage medium)
“one or more processors operably connected to the storage device and configured to execute the program instructions to cause the system to” ([0209], Computer readable program instructions can be provided to a processor for execution.)
Regarding claim 15:
Claim 15 recites an article of manufacture (“A computer program product for data autoencoding, the computer program product comprising”) that performs the same process as described in Claim 1. Therefore claim 15 is rejected under the same reasons mention for claim 1. The additional elements of claim 15 is addressed below with Yakut:
“a computer-readable storage medium having program instructions embodied thereon to perform the steps of” ([0204-0205], Instructions are stored on a computer readable storage medium)
Regarding claims 2, 9, and 16 Yakut teaches:
“wherein the Gaussian mixture comprises a combination of clusters of sub-populations of the data, wherein the clusters represent all the modalities of data” ([0114, 0146-0148], Individual modalities are Gaussian distributions and product of experts combines all individual modalities into a joint posterior (Gaussian mixture). The dataset contains 12 condition parameters (sub-population of data) and are split into two sequential modalities. The training dataset contains multiple modalities and each modality may further consist of sub-groups.)
Regarding claims 3, 10, and 17 Yakut teaches:
“wherein the clusters encode cross-modal shared information” ([0142, 0146-0148], The training dataset contains a set of data that represents the process and storage condition for a chemical production process. The dataset represents functional relationships between the two modalities, condition parameters and key performance indicators.)
Regarding claims 4, 11, and 18 Yakut teaches:
“wherein different clusters have different parameters for a same ” ([0010, 0146], Individual modalities are characterized by different statistical properties and provided to the data-driven generative model as input. The training dataset contains 12 condition parameters (different clusters have different parameters) that represents 7 years of process and storage conditions.)
Yakut does not explicitly disclose an implementation of “physics model”. However, Takeishi discloses in the same field of endeavor:
“” ([pg. 7-8, Section 5.1, par. 1-2; pg. 10, Section 5.4, par. 1-2], Different datasets containing various data was used in the experiments. The physics models are setup with different parameters to process the input data.)
It would be obvious to one of the ordinary skill in the art before the effective filing date of the claimed invention to incorporate the teaching of a physics-constraint model from Takeishi into the teaching of Yakut. Doing so can improve models with improved interpretability and abilities to extrapolate by integrating physics models into ML model (Takeishi, abstract).
Regarding claims 5, 12, and 19 Yakut teaches:
“wherein each modality of data is represented by a separate [models]” ([0114], The N individual modalities are represented by N uni-modal networks. The latent representations are passed into N independent decoder network models.)
Yakut does not explicitly disclose an implementation of “physics model”. However, Takeishi discloses in the same field of endeavor:
“wherein each modality of data is represented by a ” ([pg. 7-8, Section 5.1, par. 1-2; pg. 10, Section 5.4, par. 1-2], Different datasets containing various data was used in the experiments. The physics models are setup with different parameters to process the input data.)
It would be obvious to one of the ordinary skill in the art before the effective filing date of the claimed invention to incorporate the teaching of a physics-constraint model from Takeishi into the teaching of Yakut. Doing so can improve models with improved interpretability and abilities to extrapolate by integrating physics models into ML model (Takeishi, abstract).
Regarding claims 7, 14, and 21 Yakut teaches:
“wherein the Gaussian mixture is generated by a Product of Experts model” ([0114], The product of experts is used to combine the distributions of multiple modalities.)
Claims 6, 13, and 20 are rejected under 35 U.S.C. 103 as being unpatentable over Yakut (US20230045548A1) in view of Takeishi, “Physics-Integrated Variational Autoencoders for Robust and Interpretable Generative Modeling” and He “Deep Variational Autoencoder Classifier for Intelligent Fault Diagnosis Adaptive to Unseen Fault Categories”.
Regarding claims 6, 13, and 20 Yakut in view of Takeishi does not explicitly disclose an implementation of “wherein the encoding and decoding comprise unsupervised learning”. However, He discloses in the same field of endeavor:
“wherein the encoding and decoding comprise unsupervised learning” ([pg. 4-5, section B, par. 1 and 7], The variational autoencoder is trained using unsupervised method.)
It would be obvious to one of the ordinary skill in the art before the effective filing date of the claimed invention to incorporate the teaching of “wherein the encoding and decoding comprise unsupervised learning” from He into the teaching of Yakut in view of Takeishi. Doing so can improve the effectiveness of VAEs by training the framework with latent representations using an unsupervised method (He, abstract).
Conclusion
Any inquiry concerning this communication or earlier communications from the examiner should be directed to GARY MAC whose telephone number is (703)756-1517. The examiner can normally be reached Monday - Friday 8:00 AM - 5:00 PM.
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, Abdullah Kawsar can be reached at (571) 270-3169. 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.
/GARY MAC/Examiner, Art Unit 2127
/ABDULLAH AL KAWSAR/Supervisory Patent Examiner, Art Unit 2127