DETAILED ACTION
This action is responsive to claims filed on 19 April 2023.
Claims 9-22 are pending for examination.
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 .
Information Disclosure Statement
The information disclosure statement (IDS) submitted on 19 April 2023 is in compliance with the provisions of 37 CFR 1.97. Accordingly, the information disclosure statement is being considered and attached by the examiner.
Claim Interpretation
The following is a quotation of 35 U.S.C. 112(f):
(f) Element in Claim for a Combination. – An element in a claim for a combination may be expressed as a means or step for performing a specified function without the recital of structure, material, or acts in support thereof, and such claim shall be construed to cover the corresponding structure, material, or acts described in the specification and equivalents thereof.
The following is a quotation of pre-AIA 35 U.S.C. 112, sixth paragraph:
An element in a claim for a combination may be expressed as a means or step for performing a specified function without the recital of structure, material, or acts in support thereof, and such claim shall be construed to cover the corresponding structure, material, or acts described in the specification and equivalents thereof.
The claims in this application are given their broadest reasonable interpretation using the plain meaning of the claim language in light of the specification as it would be understood by one of ordinary skill in the art. The broadest reasonable interpretation of a claim element (also commonly referred to as a claim limitation) is limited by the description in the specification when 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, is invoked.
As explained in MPEP § 2181, subsection I, claim limitations that meet the following three-prong test will be interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph:
(A) the claim limitation uses the term “means” or “step” or a term used as a substitute for “means” that is a generic placeholder (also called a nonce term or a non-structural term having no specific structural meaning) for performing the claimed function;
(B) the term “means” or “step” or the generic placeholder is modified by functional language, typically, but not always linked by the transition word “for” (e.g., “means for”) or another linking word or phrase, such as “configured to” or “so that”; and
(C) the term “means” or “step” or the generic placeholder is not modified by sufficient structure, material, or acts for performing the claimed function.
Use of the word “means” (or “step”) in a claim with functional language creates a rebuttable presumption that the claim limitation is to be treated in accordance with 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph. The presumption that the claim limitation is interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, is rebutted when the claim limitation recites sufficient structure, material, or acts to entirely perform the recited function.
Absence of the word “means” (or “step”) in a claim creates a rebuttable presumption that the claim limitation is not to be treated in accordance with 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph. The presumption that the claim limitation is not interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, is rebutted when the claim limitation recites function without reciting sufficient structure, material or acts to entirely perform the recited function.
Claim limitations in this application that use the word “means” (or “step”) are being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, except as otherwise indicated in an Office action. Conversely, claim limitations in this application that do not use the word “means” (or “step”) are not being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, except as otherwise indicated in an Office action.
This application includes one or more claim limitations that do not use the word “means,” but are nonetheless being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, because the claim limitation(s) uses a generic placeholder that is coupled with functional language without reciting sufficient structure to perform the recited function and the generic placeholder is not preceded by a structural modifier.
Such claim limitation(s) is/are:
“setter configured to set” in claim 9.
“manager configured to convert” in claim 9.
“engine configured to input” in claim 9.
“modifier configured to cause” in claim 9.
“modifier is configured to cause” in claim 10.
“engine is configured to repeatedly change” in claim 11.
Claims 12-15, which are dependent on claim 9, are similarly interpreted.
Examiner notes, for the record, the generic placeholders listed above are listed in the Specification as embodiments of the “neural architecture search system”, which “can be implemented by a computer including a central processing unit (CPU), a storage device, and an interface and a program for controlling those hardware resources. A configuration example of the computer is illustrated in Fig. 10. The computer includes a CPU 300, a storage device 301, and an interface device (I/F) 302. The I/F 302 is connected to, for example, an external device from which information is collected. A program for implementing the neural architecture search method of embodiments of the present invention is stored in the storage device 301. The CPU 300 executes processing described in the first to fourth embodiments according to the neural architecture search program stored in the storage device 301. The program may also be provided via a network” as described in Specification [0159]-[0160].
Because this/these claim limitation(s) is/are being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, it/they is/are being interpreted to cover the corresponding structure described in the specification as performing the claimed function, and equivalents thereof.
If applicant does not intend to have this/these limitation(s) interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, applicant may: (1) amend the claim limitation(s) to avoid it/them being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph (e.g., by reciting sufficient structure to perform the claimed function); or (2) present a sufficient showing that the claim limitation(s) recite(s) sufficient structure to perform the claimed function so as to avoid it/them being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph.
Claim Rejections - 35 USC § 112
The following is a quotation of 35 U.S.C. 112(b):
(b) CONCLUSION.—The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the inventor or a joint inventor regards as the invention.
The following is a quotation of 35 U.S.C. 112 (pre-AIA ), second paragraph:
The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the applicant regards as his invention.
Claims 10, 15, 17, 22 are rejected under 35 U.S.C. 112(b) or 35 U.S.C. 112 (pre-AIA ), second paragraph, as being indefinite for failing to particularly point out and distinctly claim the subject matter which the inventor or a joint inventor (or for applications subject to pre-AIA 35 U.S.C. 112, the applicant), regards as the invention.
Claim 10 and analogous claim 17 recites the limitation "the plurality of times of learning" in line 6. There is insufficient antecedent basis for this limitation in the claim. For examination purposes, the term "the plurality of times of learning" has been construed to be “a plurality of times of learning”.
Claim 15 and analogous claim 22 recites the limitation "the processing devices" in line 5. There is insufficient antecedent basis for this limitation in the claim. For examination purposes, the term "the processing devices" has been construed to be “processing devices”.
Claim 15 and analogous claim 22 is indefinite because it is unclear whether the limitations listed in the claim “external setting parameter input by a user, communication network information that defines a constraint of a communication network that connects processing devices of the system that implements the neural network, and device information that defines a constraint of the processing devices of the system that implements the neural network” are part of “the first constraint condition” or if “wherein the first constraint condition is an external setting parameter input by a user, communication network information that defines a constraint of a communication network that connects processing devices of the system that implements the neural network, and device information that defines a constraint of the processing devices of the system that implements the neural network” listed are part of “The neural architecture search system”. For examination purposes, Claim 15 and analogous claim 22 is interpreted as “The neural architecture search system according to claim 9, wherein the first constraint condition comprises of: an external setting parameter input by a user, communication network information that defines a constraint of a communication network that connects processing devices of the system that implements the neural network, and device information that defines a constraint of the processing devices of the system that implements the neural network.”.
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.
The factual inquiries for establishing a background for determining obviousness under 35 U.S.C. 103 are summarized as follows:
1. Determining the scope and contents of the prior art.
2. Ascertaining the differences between the prior art and the claims at issue.
3. Resolving the level of ordinary skill in the pertinent art.
4. Considering objective evidence present in the application indicating obviousness or nonobviousness.
This application currently names joint inventors. In considering patentability of the claims the examiner presumes that the subject matter of the various claims was commonly owned as of the effective filing date of the claimed invention(s) absent any evidence to the contrary. Applicant is advised of the obligation under 37 CFR 1.56 to point out the inventor and effective filing dates of each claim that was not commonly owned as of the effective filing date of the later invention in order for the examiner to consider the applicability of 35 U.S.C. 102(b)(2)(C) for any potential 35 U.S.C. 102(a)(2) prior art against the later invention.
Claims 9-22 are rejected under 35 U.S.C. 103 as being unpatentable over Tan et al. (U.S. Pre-Grant Publication No. 20200143227, hereinafter ‘Tan'), in view of Denolf et al. (U.S. Pre-Grant Publication No. 20200104715, hereinafter 'Denolf').
Regarding claim 9 and analogous claim 16, Tan teaches A neural architecture search system and A neural architecture search method, respectively, comprising:
a deployment constraint manager configured to convert a first constraint condition into a second constraint condition, wherein the first constraint condition defines a constraint of a system that implements the neural network and the second constraint condition defines a constraint of a parameter that prescribes the architecture of the neural network ([0044] Referring again to example approach illustrated in FIG. 1, given a model m, let ACC (m) denote the given model's accuracy on the target task, LAT (m) denotes the inference wherein the first constraint condition defines a constraint of a system that implements the neural network latency on the target mobile platform, and T is the target latency. One possible method is to a deployment constraint manager configured to convert a first constraint condition into a second constraint condition treat T as a and the second constraint condition defines a constraint of a parameter that prescribes the architecture of the neural network hard constraint and maximize accuracy under this constraint: maximize mACC(m)subject toLAT(m)≤T(1); [0045] However, this approach only maximizes a single metric and does not provide multiple Pareto optimal solutions. Informally, a model is called Pareto optimal if either it has the highest accuracy without increasing latency or it has the lowest latency without decreasing accuracy. Given the computational cost of performing architecture search, example implementations of the present disclosure focus more on finding multiple Pareto-optimal solutions in a single architecture search.; [0046] In some implementations, the present disclosure utilizes a customized weighted product method to approximate Pareto optimal solutions, by setting the optimization goal as: maximize mACC(m)×[LAT(m)T]w(2) where w is the weight factor defined as: w=(α,ifLAT(m)≤Tβ,otherwise(3) where α and β are application-specific constants. An empirical rule for picking α and β is to check how much accuracy gain or loss is expected if the latency is doubled or halved. For example, doubling or halving the latency of MobileNetV2 brings about 5% accuracy gain or loss, so the constraints, α and β, can be empirically set α=β=−0.07, since 2−0.07−1≈1−0.5−0.07≈5%. By setting (α, β) in this way, equation (2) can effectively approximate Pareto solutions nearby the target latency T. While the weighted product method is easy to customize, other methods like weighted sum can also be used by other example implementations.; [0047] FIGS. 2A and 2B show plots of an example objective function with two typical values of (α, β). In FIG. 2A with (α=0, β=−1), accuracy is simply used as the objective value if measured latency is less than the target latency T; otherwise, the objective value is sharply penalized to discourage models from violating latency constraints. In FIG. 2B (α=β=−0.07) the objective function treats the target latency T as a soft constraint, and smoothly adjusts the objective value based on the measured latency. In some implementations of the present disclosure, the application-specific constraints, α and β, are set to α=/β=−0.07 in order to obtain multiple Pareto optimal models in a single search experiment. In some implementations, reward functions may dynamically adapt to the Pareto curve.);
a learning engine configured to input training data to the neural network, perform learning of the neural network under the search condition, and calculate inference accuracy in a case where inference is performed by using the learned neural network ([0035] Thus, at each of a plurality of iterations, the search system can modify at least one of the searchable parameters in the sub-search space associated with at least one of the plurality of blocks to generate one or more new network structures for an artificial neural network. For example, the modifications can be guided by a controller (e.g., a recurrent neural network-based controller) or can be random (e.g., random evolutionary mutations).; [0036] In some implementations, the search system can measure one or more performance characteristics of the new network structures for the artificial neural network. The search system can use the measured performance characteristics to, for example, determine whether to keep or discard the new network structure (e.g., through comparison to performance characteristics of a best-previously-observed structure). Additionally or alternatively, the search system can use the measured performance characteristics to determine a reward to provide to the controller in a reinforcement learning scheme and/or other measurements of loss, reward, regret, and/or the like (e.g., for use in gradient-based optimization schemes). As an example, the calculate inference accuracy in a case where inference is performed by using the learned neural network measured performance characteristics can include an accuracy (or an estimated accuracy) of the perform learning of the neural network under the search condition network structure as trained for and evaluated on a learning engine configured to input training data to the neural network a particular training dataset and/or prediction task.); and
a model modifier configured to cause the learning engine to repeatedly perform the learning and the calculating of the inference accuracy while changing the architecture of the neural network based on the inference accuracy and the second constraint condition so as to obtain an architecture having a best inference accuracy ([0037] According to another aspect, in some implementations, the measured performance characteristics can include a real-world latency associated with implementation of the new network structure on a real-world mobile device. More particularly, in some implementations, the search system can explicitly incorporate latency information into the main objective so that the search can identify a model that achieves a good trade-off between accuracy and latency.; [0052] Referring again to FIG. 1, one example search framework consists of three components: a controller (e.g., a recurrent neural network (RNN) based controller), a trainer to obtain the model accuracy, and a mobile phone-based inference engine for measuring the latency. The a model modifier configured to cause the learning engine to repeatedly perform the learning and the calculating of the inference accuracy framework can use a sample-eval-update loop to train the controller. At each step, the controller first samples a batch of models using its current parameters θ, (e.g., by predicting a sequence of tokens based on the softmax logits from its RNN). For each sampled model m, it is trained on the target task to get its accuracy ACC (m), and run on real phones to get its inference latency LAT (m). Then, the reward value R(m) is calculated using equation (2). At the while changing the architecture of the neural network end of each step, the parameters θ of the controller are updated by maximizing the based on the inference accuracy and the second constraint condition so as to obtain an architecture having a best inference accuracy expected reward defined by equation (4) (e.g., using Proximal Policy Optimization (Schulman et al. 2017)). The sample-eval-update loop can be repeated until it reaches the maximum number of steps or the parameters θ converge.).
Tan fails to teach a search parameter setter configured to set a search condition of an architecture of a neural network;
Denolf teaches a search parameter setter configured to set a search condition of an architecture of a neural network ([0025] The implementation efficiency of a neural network implementation can be measured by different costs, such as throughput, energy, size, error tolerance, and the like, or combinations thereof. This a search parameter setter configured to set cost is the result of different design aspects, such as the a search condition of an architecture of a neural network number of operations, bandwidth, data locality, scheduling on the hardware backend, and the like. These aspects are related to the characteristics of the training algorithm, where a better algorithmic performance often leads to higher implementation costs (Pareto principle). Typically, maximizing the algorithmic accuracy for a specific task/capability is the main objective during training.);
Tan and Denolf are considered to be analogous to the claimed invention because they are in the same field of machine learning. In view of the teachings of Tan, it would have been obvious for a person of ordinary skill in the art to apply the teachings of Denolf to Tan before the effective filing date of the claimed invention in order to balance accuracy against the implementation cost of the neural network during neural network training (cf. Denolf, [0022] Techniques for training of neural network by including implementation cost as an objective are described. The techniques provide a cost-aware architectural search of a neural network topology. As such, the training of a neural network no longer only targets maximizing the accuracy of the neural network at a certain task. Rather, the neural network training balances accuracy against the implementation cost of the neural network, which is included as another objective in the training. In this manner, the training becomes a multi-objective search, where not only the values of the weights are trained, but also the topology and certain implementation-related attributes of the neural network are found.).
Regarding claim 10 and analogous claim 17, Tan, as modified by Denolf, teaches The neural architecture search system of claim 9 and The neural architecture search method of claim 16, respectively.
Tan teaches wherein the model modifier is configured to cause the learning engine to perform the learning and the calculating of the inference accuracy while changing the architecture of the neural network a plurality of times so as to satisfy the second constraint condition and to obtain, as a final search result, the architecture having the best inference accuracy among a plurality of architectures of the neural network obtained by the plurality of times of learning ([0052]
Referring again to FIG. 1, one example search framework consists of three components: a controller (e.g., a recurrent neural network (RNN) based controller), a trainer to obtain the model accuracy, and a mobile phone-based inference engine for measuring the latency. The framework can use a sample-eval-update loop to train the controller. At each step, the controller first samples a batch of models using its current parameters θ, (e.g., by predicting a sequence of tokens based on the softmax logits from its RNN). while changing the architecture of the neural network a plurality of times For each sampled model m, it is trained on the target task to get its to satisfy the second constraint condition accuracy ACC (m), and run on real phones to get its inference latency LAT (m). Then, the model modifier is configured to cause the learning engine to perform the learning and the calculating of the inference accuracy reward value R(m) is calculated using equation (2). At the end of each step, the parameters θ of the controller are updated by maximizing the expected reward defined by equation (4) (e.g., using Proximal Policy Optimization (Schulman et al. 2017)). The sample-eval-update loop can be repeated until it reaches the maximum number of steps or the parameters θ converge.; [0128] In some implementations, obtain, as a final search result, the architecture having the best inference accuracy among a plurality of architectures of the neural network obtained by the plurality of times of learning determining the outcome at 910 can include comparing the performance characteristics of the new network structure to those of a best-previously-observed structure to determine whether to keep or discard the new network structure.; [0129] After 910, the method 900 can optionally return to block 906 and again determine a new network structure. Thus, blocks 906-910 can be iteratively performed to identify and evaluate new network structures.).
Tan and Denolf are combinable for the same rationale as set forth above with respect to claim 9.
Regarding claim 11 and analogous claim 18, Tan, as modified by Denolf, teaches The neural architecture search system of claim 9 and The neural architecture search method of claim 16, respectively.
Tan teaches wherein the learning engine is configured to repeatedly change the architecture of the neural network and perform the learning under the search condition and to obtain the architecture having the best inference accuracy as a learning result ([0035] Thus, at each of a learning engine is configured to repeatedly change the architecture of the neural network and perform the learning under the search condition plurality of iterations, the search system can modify at least one of the searchable parameters in the sub-search space associated with at least one of the plurality of blocks to to obtain the architecture having the best inference accuracy as a learning result generate one or more new network structures for an artificial neural network. For example, the modifications can be guided by a controller (e.g., a recurrent neural network-based controller) or can be random (e.g., random evolutionary mutations).).
Tan and Denolf are combinable for the same rationale as set forth above with respect to claim 9.
Regarding claim 12 and analogous claim 19, Tan, as modified by Denolf, teaches The neural architecture search system of claim 9 and The neural architecture search method of claim 16, respectively.
Tan teaches wherein the first constraint condition is an external setting parameter from a user ([0110] FIG. 8A illustrates one example computing system that can be used to implement the present disclosure. Other computing systems can be used as well. For example, in some implementations, the user computing device 102 can include the model trainer 160 and the training dataset 162. In such implementations, the networks 120 can be both trained and used locally at the user computing device 102. In some of such implementations, the user computing device 102 can implement the model trainer 160 to wherein the first constraint condition is an external setting parameter from a user personalize the networks 120 based on user-specific data.).
Tan and Denolf are combinable for the same rationale as set forth above with respect to claim 9.
Regarding claim 13 and analogous claim 20, Tan, as modified by Denolf, teaches The neural architecture search system of claim 9 and The neural architecture search method of claim 16, respectively.
Denolf teaches wherein the first constraint condition is communication network information that defines a constraint of a communication network that connects processing devices of the system that implements the neural network ([0038] A general formulation of multi-objective optimization is as follows: minx(f1(x),f2(x),…,fk(x))s.t.x∈X, where f1, . . . , fx are functions that define the cost of each objective that is being optimized, x is a vector representing the current solution, and X is the search space of all possible solutions. In the examples described herein, x represents a neural network topology and its associated hyperparameters (i.e., the model-capacity hyperparameters 108). The functions represent metrics of interest of the current neural network topology in relation to its accuracy and implementation/hardware cost. For accuracy, these functions include mean squares error (MSE), classification error, lp norm, hingle loss, or a similar metric suitable for the target domain. For wherein the first constraint condition implementation/hardware cost, these functions include memory requirements, is communication network information that defines a constraint of a communication network that connects processing devices of the system that implements the neural network bandwidth requirements, clock cycles, datapath width, quantization scheme, arithmetic style, number formats, silicon area, and energy consumption, and error tolerance.).
Tan and Denolf are combinable for the same rationale as set forth above with respect to claim 9.
Regarding claim 14 and analogous claim 21, Tan, as modified by Denolf, teaches The neural architecture search system of claim 9 and The neural architecture search method of claim 16, respectively.
Tan teaches wherein the first constraint condition is device information that defines a constraint of processing devices of the system that implements the neural network ([0037]
According to another aspect, in some implementations, the measured performance characteristics can include a real-world latency associated with implementation of the new network structure on a real-world mobile device. More particularly, in some implementations, the search system can explicitly incorporate latency information into the main objective so that the search can identify a model that achieves a good trade-off between accuracy and latency. Unlike in previous work, where mobile latency is considered via another, often inaccurate proxy (e.g., FLOPS), in some implementations, first constraint condition is device information that defines a constraint of processing devices of the system that implements the neural network real-world inference latency can be directly measured by executing the model on a particular platform (e.g., a mobile device such as the Google Pixel device). In further implementations, various other performance characteristics can be included in a multi-objective function that guides the search process, including, as examples, power consumption, user interface responsiveness, peak compute requirements, and/or other characteristics of the generated network structures.).
Tan and Denolf are combinable for the same rationale as set forth above with respect to claim 9.
Regarding claim 15 and analogous claim 22, Tan, as modified by Denolf, teaches The neural architecture search system of claim 9 and The neural architecture search method of claim 16, respectively.
Tan teaches wherein the first constraint condition is an external setting parameter input by a user ([0110] FIG. 8A illustrates one example computing system that can be used to implement the present disclosure. Other computing systems can be used as well. For example, in some implementations, the user computing device 102 can include the model trainer 160 and the training dataset 162. In such implementations, the networks 120 can be both trained and used locally at the user computing device 102. In some of such implementations, the user computing device 102 can implement the model trainer 160 to wherein the first constraint condition is an external setting parameter input by a user personalize the networks 120 based on user-specific data.),
device information that defines a constraint of the processing devices of the system that implements the neural network ([0037] According to another aspect, in some implementations, the measured performance characteristics can include a real-world latency associated with implementation of the new network structure on a real-world mobile device. More particularly, in some implementations, the search system can explicitly incorporate latency information into the main objective so that the search can identify a model that achieves a good trade-off between accuracy and latency. Unlike in previous work, where mobile latency is considered via another, often inaccurate proxy (e.g., FLOPS), in some implementations, first constraint condition is device information that defines a constraint of processing devices of the system that implements the neural network real-world inference latency can be directly measured by executing the model on a particular platform (e.g., a mobile device such as the Google Pixel device). In further implementations, various other performance characteristics can be included in a multi-objective function that guides the search process, including, as examples, power consumption, user interface responsiveness, peak compute requirements, and/or other characteristics of the generated network structures.).
Denolf teaches communication network information that defines a constraint of a communication network that connects processing devices of the system that implements the neural network ([0038] A general formulation of multi-objective optimization is as follows: minx(f1(x),f2(x),…,fk(x))s.t.x∈X, where f1, . . . , fx are functions that define the cost of each objective that is being optimized, x is a vector representing the current solution, and X is the search space of all possible solutions. In the examples described herein, x represents a neural network topology and its associated hyperparameters (i.e., the model-capacity hyperparameters 108). The functions represent metrics of interest of the current neural network topology in relation to its accuracy and implementation/hardware cost. For accuracy, these functions include mean squares error (MSE), classification error, lp norm, hingle loss, or a similar metric suitable for the target domain. For wherein the first constraint condition implementation/hardware cost, these functions include memory requirements, is communication network information that defines a constraint of a communication network that connects processing devices of the system that implements the neural network bandwidth requirements, clock cycles, datapath width, quantization scheme, arithmetic style, number formats, silicon area, and energy consumption, and error tolerance.), and
Tan and Denolf are combinable for the same rationale as set forth above with respect to claim 9.
Conclusion
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure.
Lu et al. (NPL: “On Neural Architecture Search for Resource-Constrained Hardware Platforms”) teaches a new framework to jointly explore the space of neural architecture, hardware implementation, and quantization.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to MAGGIE MAIDO whose telephone number is (703) 756-1953. The examiner can normally be reached M-Th: 6am - 4pm.
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, Michael Huntley can be reached on (303) 297-4307. 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.
/MM/Examiner, Art Unit 2129
/MICHAEL J HUNTLEY/Supervisory Patent Examiner, Art Unit 2129