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 .
This action is in response to the application filed on 11/29/2022. Claims 1-15 are pending in the case.
Information Disclosure Statement
As required by MPEP 609 (c), the Applicants’ submission of the Information Disclosure Statement(s) filed on 07/17/2023 are acknowledged by the examiner and the cited references have been considered in the examination of the claims now pending.
Claim Rejections - 35 USC § 101
5. 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 1-20 rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more.
Step 1
According to the first part of the analysis, in the instant case, claims 1-9 are directed to a method and claims 10-15 are directed to a system claim. Thus, each of the claims falls within one of the four statutory categories (i.e., process, machine, manufacture, or composition of matter).
Regarding Claim 1
At step 2A, prong 1, Does the claim recite a judicial exception?
Independent claims 1, and 9, the steps of:
selecting, in batches, at least two collaborators from the multiple collaborators (This step involves selecting a data source and is understood to be a mental process )
receiving corresponding semantic segmentation model information from each of the at least two collaborators (This step involves obtaining information and is understood to be a mental process )
assigning respective weight parameters to the semantic segmentation model information from each of the at least two collaborators ((This step involves mathematical manipulation and is understood to be a recitation of a mathematical operations),
adding controlled noise to the weight parameters for the semantic segmentation model information from each of the at least two collaborators (This step involves mathematical operation and is understood to be a recitation of a mathematical operations), and
generating an aggregated model based on the semantic segmentation model information using the weight parameters with the noise added thereto from each of the at least two collaborators (This step involves mathematical modeling and data processing to produce new model form existing model and is understood to be a recitation of a mental process or mathematical concept)
The recites a process that can be carried out using mathematical operation and information processing . The claim recite under broadest reasonable interpretation, involves mathematical manipulation of information by applying data processing and model aggregation. Because the claims involve mainly mathematical operations on collected information the claim falls within the “Mathematical concepts” groupings of abstract ideas. Accordingly, the claims recite an abstract idea.
Step 2A prong 2: Does the claim recite additional elements? Do those additional elements, individually and in combination, integrate the judicial exception into a practical application?
Further, the claim does not recite any additional element which could integrate this abstract idea into a practical application, because the additional elements recited of consist of:
semantic segmentation models (claim1), an aggregator server (claim 9), a communication network (claim 9). These steps describe mere instructions to apply the exception using generic computer components - see MPEP 2106.05(f). The additional elements as disclosed above alone or in combination do not integrate the judicial exception into practical application. Therefore, these limitations, taken alone or in combination, do not integrate the abstract idea into a practical application.
Step 2B: Do the additional elements, considered individually and in combination, amount to significantly more than the judicial exception?
No, as shown above with respect to integration of the abstract idea into a practical application, the additional element of “semantic segmentation models (claim1), an aggregator server (claim 9), a communication network (claim 9), does not provide “significantly more” than performing math on data. Mere instructions to apply an exception using generic computer components cannot provide an inventive concept. Neither can insignificant extra-solution activity. All of these additional elements as generically claimed as they are insignificant extra solution activity in combination of generic computer functions.
Therefore, these limitations, taken alone or in combination, do not integrate the abstract idea into a practical application or recite significantly more that the abstract idea.
Thus, these independent claims are not patent eligible.
The dependent claims respectively recite a judicial exception in limitations of: “implementing a sliding window technique over a randomized index of the multiple collaborators for selecting, in batches, the at least two collaborators from the multiple collaborators.”(claims 2/10), “assigning higher weight parameter to the semantic segmentation model information being closer to a non-weighted average of the semantic segmentation model information from each of the at least two collaborators in comparison to the semantic segmentation model information of other of the at least two collaborators.”(claims 3/11), “assigning higher weight parameter to the semantic segmentation model information based on a corresponding database size being larger in comparison to the semantic segmentation model information of other of the at least two collaborators (claims 4/12), “the added noise is based on at least one of: a privacy budget parameter ɛ, a sensitivity scaling factor α, and a database size of the corresponding semantic segmentation model information.” (claim 5/13), “the semantic segmentation model information from each of the at least two collaborators is received at an aggregator server and wherein the aggregated model is generated at the aggregator server.”(claims 6), “transferring the aggregated model by the aggregator server to each of the multiple collaborators.”(claims 7/14), “optimizing the receiving of the semantic segmentation model information from each of the at least two collaborators at the aggregator server and the transferring of the aggregated model by the aggregator server to each of the multiple collaborators by utilizing at least one of: sparsity transformations algorithm, pruning algorithm, or encoded polyline utility algorithm.”(Claim 8/15).
These additional limitations (in claims 2-8 and 10-15) also constitute concepts performed in the human mind which fall within the “Mathematical Concept” groupings of abstract ideas.
This judicial exception is not integrated into a practical application. Additional elements “computer readable medium comprising: computer program code (in claims 2-8 and 10-15) all amount to no more than adding insignificant extra-solution activity/specifications related to data gathering, data input, or data transmittal. These additional elements do not integrate the abstract idea into a practical application because they do not impose any meaningful limits on practicing the abstract idea. The dependent claims do not include additional elements that are sufficient to amount to significantly more than the judicial exception. As discussed above with respect to integration of the abstract idea into a practical application, the additional elements of non-transitory computer readable medium comprising: computer program code are again insignificant extra-solution activity steps that cannot provide an inventive concept. All of these additional elements as generically claimed are considered well-understood, routine, and conventional.
Therefore, these limitations, taken alone or in combination, do not integrate the abstract idea into a practical application or recite significantly more that the abstract idea. Thus, all of the dependent claims are also not patent eligible.
Examiner Comments
4. In the event the determination of the status of the application as subject to AIA 35 U.S.C. 102 and 103 (or as subject to pre-AIA 35 U.S.C. 102 and 103) is incorrect, any correction of the statutory basis for the rejection will not be considered a new ground of rejection if the prior art relied upon, and the rationale supporting the rejection, would be the same under either status.
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 1-15 are rejected under 35 U.S.C. 103 as being unpatentable over Balakrishnan (US 20230177349 A1 : Pub. Date: 2023-06-08) in view of Peterson (US 20210073677 A1, Date Published 2021-03-11) in further view of Jian ( NPL: Semantic Image Segmentation With Propagating Deep Aggregation, PUB. 12/2020)
Regarding independent Claim 1,
Balakrishnan teaches a method for aggregation of semantic segmentation (see Balakrishnan teaching process for performing federated meta-learning using a clustering algorithm), the method comprising:
selecting, in batches, at least two collaborators from the multiple collaborators (see Balakrishnan: Fig.1, [0138], “shown, at 1302, global model weights (e.g., weights for model 1204 of FIG. 12) are sent by a central server (e.g., 1208 of FIG. 12) to a selected set of clients (e.g., 1202 of FIG. 12). The clients may be selected by the central server in any suitable way. For example, in some instances, K clients may be selected randomly from N total clients. As another example, the clients may be clustered (by the central server or by another edge compute node), and clients may be selected from the clusters.”)
receiving corresponding semantic segmentation model information from each of the at least two collaborators (see Balakrishnan: Fig.1, [0138], “shown, at 1302, global model weights (e.g., weights for model 1204 of FIG. 12) are sent by a central server (e.g., 1208 of FIG. 12) to a selected set of clients (e.g., 1202 of FIG. 12). The clients may be selected by the central server in any suitable way. For example, in some instances, K clients may be selected randomly from N total clients. As another example, the clients may be clustered (by the central server or by another edge compute node), and clients may be selected from the clusters.”) , [0158], “At 1510, each of the K clients sends the parameterized weight updates computed at 1506 (w.sub.t+1.sup.k) and Hessians computed at 1508 (h.sub.k) to the central server.”, i.e., the server receives weights gradients parameters form the selected clients )
assigning respective weight parameters to the semantic segmentation model information from each of the at least two collaborators (see Balakrishnan: Fig.15, [0158], “At 1510, each of the K clients sends the parameterized weight updates computed at 1506 (w.sub.t+1.sup.k) and Hessians computed at 1508 (h.sub.k) to the central server.” … [0159], “At 1512, the central server utilizes the gradient expression for the ML model and evaluates the gradient at each of the local weight updates as g.sub.k(ŵ.sub.t+1.sup.k) using a sample dataset corresponding to each client D.sub.k.sup.test.”));
generating an aggregated model based on the semantic segmentation model information using the weight parameters (see Balakrishnan: Fig.15, [0158], “At 1508, each client computes its Hessian h.sub.k on the gradient g.sub.k computed at 1504. The Hessian may be computed on the local dataset D.sub.in, or in some instances, may be computed on a separate dataset, as long as the separate dataset is i.i.d. drawn from q.sub.k.”)
Balakrishnan does not teach the system wherein:
adding controlled noise to the weight parameters for the semantic segmentation model information from each of the at least two collaborators
generating an aggregated model based on the semantic segmentation model information using the weight parameters with the noise added thereto from each of the at least two collaborators.
However, Peterson teaches the method wherein:
adding controlled noise to the weight parameters for the semantic segmentation model information from each of the at least two collaborators (see Peterson: Fig.2, [0060], “The noise should be random to thwart reverse engineering by a spy. For example, non-random noise may have a detectable pattern, for which a spy may compensate. Step 204 adds random noise 170 to gradient 160B to generate noisy gradient 180.” … [0159], “At 1512, the central server utilizes the gradient expression for the ML model and evaluates the gradient at each of the local weight updates as g.sub.k(ŵ.sub.t+1.sup.k) using a sample dataset corresponding to each client D.sub.k.sup.test.”);
generating an aggregated model based on the semantic segmentation model information using the weight parameters with the noise added thereto from each of the at least two collaborators (see Peterson: Fig.2,( [0070], “In step 207, central server 120 adjusts general ML model 130A based on noisy gradients sent from client devices 111-112 in step 205. As independent agents, the client device steps 202-206 may occur at different times for different client devices. Depending on the embodiment, even client device step 201 may occur at different times for different client devices, such as if client device 111 polls or schedules to pull coefficients 151A and 152B from central server 120. Pull frequency of different client devices may be different, or may be the same but out of phase.”)
Because both Balakrishnan and Peterson are in the same/similar field of endeavor of federated learning by the ML model, accordingly, it would have been obvious to a person of ordinary skill in the art, before the effective filing date of the invention, to modify the teaching of Balakrishnan to include the method that add controlled noise to the weight parameters for the semantic segmentation model information and generate generating an aggregated model using the weight parameters with the noise added as taught by Peterson. After modification of Balakrishnan, the federated learning that batch participants and weighing updates can also incorporate the method of adding noise to federated aggregation to improve privacy and efficiency of a model as taught Peterson. One would have been motivated to make such a combination in order to increase privacy (for example by transferring results instead of raw data), to automate and scale ML training, to exploit wireless for computation including over the air combining, and to promote multi-stage learning.( see Peterson [0028]
Examiner notes that semantic segmentation model is applying known technique (federated segregation) to particular type of neural network task (semantic segmentation) and it is treated as field -of use limitation. In order to provide compact prosecution , the following NPL is used to teach federated semantic segmentation model)
Jian teaches aggregation of semantic segmentation model (see: Jian: Abstract: we propose a new architecture of feature aggregation, which is designed to deal with the problem that the information of each convolutional layer cannot be used reasonably and the shallow layer information is lost in the process of transmission. In this method, propagating deep aggregation (PDA) is transplanted into the Deep Lab-ASPP model to generate a new model combining structural continuity with feature aggregation.”)
It would have been obvious to a person of ordinary skill in the art, before the effective filing date of the invention, to modify the teaching of Balakrishnan and to include semantic segmentation model information and generate generating an aggregated model using the weight parameters with the noise added as taught by Jian. One would have been motivated to make such a combination in order to provide efficient and secure and privacy federated learning application.
Regarding Claim 2,
As shown above, Balakrishnan, Peterson and Jian teaches all the limitations of Claim 1. Balakrishnan further teaches the method comprising:
implementing a sliding window technique over a randomized index of the multiple collaborators for selecting, in batches, the at least two collaborators from the multiple collaborators (see Balakrishnan: Fig.15, [0190], “To determine each client's batch size, an estimation of the clients' compute time as a function of the number of training examples may be determined. This may be determined by the clients themselves and transmitted to the server (a client-based approach), or may be estimated by the server based on information sent from the clients to the server (a server-based approach). A client with a large number of data points may utilize just a subset of its dataset during training in order to meet a reference time duration T.sub.ref, which may refer to a particular amount of time in which clients may be assigned to perform an update.”)
Regarding Claim 3,
As shown above, Balakrishnan, Peterson and Jian teaches all the limitations of Claim 1. Balakrishnan further teaches the method comprising:
assigning higher weight parameter to the semantic segmentation model information being closer to a non-weighted average of the semantic segmentation model information from each of the at least two collaborators in comparison to the semantic segmentation model information of other of the at least two collaborators (see Balakrishnan: Fig.15, “each client computing node 1202 fetches or otherwise obtains a global model 1204 from a central server 1208 (e.g., a MEC server) coupled to an access point 1210 (e.g., a base station), updates aspects of the global model (e.g., model parameters or weights used in the global model, e.g., NN node weights) using its local data or data provided by the central server (e.g., a subset of a large training dataset D), and communicates the updates to the global model to the central server 1208. The central server 1208 then aggregates (e.g., averages) the received updates and obtains a final global model based on the aggregated updates (e.g., updates the model weight values based on an average of the weight values received from the clients). Federated learning may be more efficient than asynchronous update methods as it avoids the prohibitive number of model updates both at the central server and worker computing nodes.”)
Regarding Claim 4,
As shown above Balakrishnan, Peterson and Jian teaches all the limitations of Claim 1. Balakrishnan further teaches the method comprising:
assigning higher weight parameter to the semantic segmentation model information based on a corresponding database size being larger in comparison to the semantic segmentation model information of other of the at least two collaborators (see Balakrishnan: Fig.15, [0190], “To determine each client's batch size, an estimation of the clients' compute time as a function of the number of training examples may be determined. This may be determined by the clients themselves and transmitted to the server (a client-based approach), or may be estimated by the server based on information sent from the clients to the server (a server-based approach). A client with a large number of data points may utilize just a subset of its dataset during training in order to meet a reference time duration T.sub.ref, which may refer to a particular amount of time in which clients may be assigned to perform an update. The batch size selected for each client may be based on the maximum amount of data the client may use in computing an update such that the update may be performed within the time indicated by T.sub.ref.”)
Regarding Claim 5,
As shown above, Balakrishnan, Peterson and Jian teaches all the limitations of Claim 1. Balakrishnan further teaches the method comprising:
the added noise is based on at least one of: a privacy budget parameter ɛ, a sensitivity scaling factor α, and a database size of the corresponding semantic segmentation model information (see Peterson: Fig.2, [0008]“Informally, differential privacy ensures privacy of training data by introducing “noise” in the training process (inputs, parameters, or outputs), in order to bound the variation in the output, to a predefined quantity ε (epsilon), based on inclusion/exclusion of a single data point from a given data set.”)
Regarding Claim 6,
As shown above, Balakrishnan, Peterson and Jian teaches all the limitations of Claim 1. Balakrishnan further teaches the method comprising:
the semantic segmentation model information from each of the at least two collaborators is received at an aggregator server and wherein the aggregated model is generated at the aggregator server (see Balakrishnan: [0154], Fig.15, “At 1502, the central server selects a set of K clients from a number N clients. In some instances, the central server draws the set K of clients uniformly from the distribution p of N clients, where each client has its own underlying data distribution q.sub.k and the data (x,y) represent a d-dimensional training data and label respectively. In other instances, the central server selects the K clients based on the clustering approach described further below. The central server sends global model weights w.sub.t to the selected K clients.”)
Regarding Claim 7,
As shown above, Balakrishnan, Peterson and Jian teaches all the limitations of Claim 1. Balakrishnan further teaches the method comprising:
transferring the aggregated model by the aggregator server to each of the multiple collaborators (see Balakrishnan: Fig.15, [0154], “At 1502, the central server selects a set of K clients from a number N clients. In some instances, the central server draws the set K of clients uniformly from the distribution p of N clients, where each client has its own underlying data distribution q.sub.k and the data (x,y) represent a d-dimensional training data and label respectively. In other instances, the central server selects the K clients based on the clustering approach described further below. The central server sends global model weights w.sub.t to the selected K clients.”)
Regarding Claim 8,
As shown above, Balakrishnan, Peterson and Jian teaches all the limitations of Claim 6. Balakrishnan further teaches the method comprising:
optimizing the receiving of the semantic segmentation model information from each of the at least two collaborators at the aggregator server (see Balakrishnan: Fig.15 [0159], “At 1512, the central server utilizes the gradient expression for the ML model and evaluates the gradient at each of the local weight updates as g.sub.k(ŵ.sub.t+1.sup.k) using a sample dataset corresponding to each client D.sub.k.sup.test.”), and the transferring of the aggregated model by the aggregator server to each of the multiple collaborators by utilizing at least one of: sparsity transformations algorithm, pruning algorithm, or encoded polyline utility algorithm (see Balakrishnan: Fig.15, [0213], “machine learning (ML) approach, specifically, an end-to-end deep reinforcement learning (RL), that allows a central server to select clients to participate in an aggregate gradient update for each training epoch. The proposed methods herein allow agents to interact with the environment (e.g., a system similar to system 1200 of FIG. 12) and take actions that increasingly allow for maximization of the expected long-term rewards.”)
Regarding independent Claim 9,
Claim 9 is directed to a system claim and has similar/same claim limitations as Claim 1 and is rejected under the same rationale.
Regarding Claim 10,
Claim 10 is directed to a system claim and has similar/same claim limitations as Claim 2 and is rejected under the same rationale.
Regarding Claim 11,
Claim 11 is directed to a system claim and has similar/same claim limitations as Claim 3 and is rejected under the same rationale.
Regarding Claim 12
Claim 12 is directed to a system claim and has similar/same claim limitations as Claim 4 and is rejected under the same rationale.
Regarding Claim 13,
Claim 13 is directed to a system claim and has similar/same claim limitations as Claim 5 and is rejected under the same rationale.
Regarding Claim 14,
Claim 14 is directed to a system claim and has similar/same claim limitations as Claim 7 nd is rejected under the same rationale.
Regarding Claim 15,
Claim 15 is directed to a system claim and has similar/same claim limitations as Claim 8 and is rejected under the same rationale.
Conclusion
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure.
PGPUB
NUMBER:
INVENTOR-INFORMATION:
TITLE / DESCRIPTION
US 20240104367 A1
LIN; Jamie Menjay
Title: MODEL DECORRELATION AND SUBSPACING FOR FEDERATED LEARNING
Description: Federated learning generally refers to various techniques that allow for training a machine learning model to be distributed across a plurality of client devices, which beneficially allows for a machine learning model to be trained using a wide variety of data.
US 20230394365 A1
GUO; Hua
Title: FEDERATED LEARNING PARTICIPANT SELECTION METHOD AND APPARATUS, AND DEVICE AND STORAGE MEDIUM
Description: The present disclosure relates to the technical field of communication, in particular to a method, a device, and an apparatus for selecting participants, and a storage medium.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to ZELALEM W SHALU whose telephone number is (571)272-3003. The examiner can normally be reached M- F 0800am- 0500pm.
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, Cesar Paula can be reached at (571) 272-4128. 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.
Zelalem Shalu
Examiner
Art Unit 2145
/Zelalem Shalu/Examiner, Art Unit 2145
/CESAR B PAULA/Supervisory Patent Examiner, Art Unit 2145