The present application, filed on or after 16 March 2013, is being examined under the first inventor to file provisions of the AIA .
DETAILED ACTION
This office action is in response to Applicant’s submission filed on 31 August 2020. THIS ACTION IS NON-FINAL.
Status of Claims
Claims 1-15 are pending.
Claims 9-15 are withdrawn
Claim 1-2, 4-8 are rejected under 35 U.S.C. 101 for being directed to a judicial exception (i.e., a law of nature, a natural phenomenon, or an abstract idea) without significantly more.
Claim 3 is objected to.
There is no art rejection for claims 1-8.
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.
Judicial Exception
Claims 1-2, 4-8 of the claimed invention are directed to a judicial exception, an abstract idea, without significantly more.
(Independent Claims) With regards to claim 1, the claim recites a process, which falls into one of the statutory categories.
2A – Prong 1: Claim 1, in part, recites
“… setting a condition of an objective function on the basis of combinations of pruning rates respectively applied to the plurality of convolutional layers, wherein the condition is that specifies the combination of pruning rates minimizing a value of the objective function minimizes a difference between filters of the plurality of convolutional layers and filters of the plurality of convolutional layers pruned by the combination of pruning rates minimizing the value of the objective function; and
determining the combination of pruning rates minimizing the value of the objective function as a combination of optimal pruning rates from the objective function on the basis of Bayesian optimization,
wherein the objective function is defined as a sum of weight change rates across all of the plurality of convolutional layers, the weight change rates resulting from pruning each of the plurality of convolutional layers, and
wherein the conditions of the objective function are that (i) a storage cost ratio between a filter before pruninq and a filter after pruninq is equal to or less than a first threshold value, and (ii) a computational cost ratio between the filter before pruning and the filter after pruning is equal to or less than a second threshold value” (mental process and/or math concept), as drafted, is a process that, under its broadest reasonable interpretation, covers performance of the limitation in the mind but for the recitation of generic computer components. That is, other than reciting a computing device, nothing in the claim element precludes the step from practically being performed in the mind. For example, but for the language about generic computer components, “setting”, “determining”, in the limitation citied above encompasses setting condition and evaluating based on certain math function to determine parameters for optimizing an abstract data processing model (neural network) which is based on observation, evaluation, judgement, and/or opinion, that could be performed by human using paper / pen / calculator. If a claim limitation, under its broadest reasonable interpretation, covers performance of the limitation in the mind but for the recitation of generic computer components, then it falls within the “Mental Processes” grouping of abstract ideas. Accordingly, the claim recites an abstract idea.
2A – Prong 2: This judicial exception is not integrated into a practical application. In particular, the claim recites the additional elements: “acquiring the target neural network comprising the plurality of convolutional layers“ (insignificant extra solution activity (MPEP.2106.05(g)) and/or WURC (MPEP 2106/05(d)(II))), which are recited at a high level of generality and amounts to extra-solution activity of receiving data i.e. pre-solution activity of gathering data for use in the claimed process as described in MPEP.2106.05(g). The courts have found limitations directed to obtaining information electronically, recited at a high level of generality, to be well-understood, routine, and conventional (see MPEP 2106.05(d)(II), “receiving or transmitting data over a network”, "electronic record keeping," and "storing and retrieving information in memory"). There is no additional elements showing integration of the abstract idea into a practical application and/or providing anything significantly more to the abstract idea. The claim is directed to an abstract idea.
2B Analysis: The claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception. As discussed above, the additional element of generic computer element merely uses generic computer as a tool to perform the abstract idea (MPEP 2106.05(f)). The additional element of “acquiring the target neural network comprising the plurality of convolutional layers”, is insignificant extra-solution activity of data input / output (MPEP 2106.05(f)). The courts have found limitations directed to obtaining information electronically, recited at a high level of generality, to be well-understood, routine, and conventional (MPEP 2106.05(d)(II)), “receiving or transmitting data over a network”, "electronic record keeping," and "storing and retrieving information in memory"). Hence the additional elements do not add anything significant to the abstract idea. The claim is not patent eligible.
(Dependent claims)
Claims 2-8 are dependent on claim 1 and include all the limitations of claim 1. Therefore, claims 2, 4-8 recite the same abstract ideas.
With regards to claim 2, the claim recites “wherein the target neural network further comprises a plurality of pooling layers corresponding to the plurality of convolutional layers, and at least one full connection layer” which is further limitation on neural network model for evaluation, which is still part of an abstract idea. The claim does not include additional elements that shows integration into a practical application or adding something significantly more to the judicial exception. The claim is not patent eligible.
With regards to claim 4, the claim recites “wherein the target neural network is a convolutional neural network” which is further limitation on neural network model for evaluation, which is still part of an abstract idea. The claim does not include additional elements that shows integration into a practical application or adding something significantly more to the judicial exception. The claim is not patent eligible.
With regards to claim 5, the claim recites “wherein the determining of the combination of optimal pruning rates comprises performing, when a dimension of the objective function is equal to or greater than a threshold value, projection on the objective function into a dimension less than the threshold value and applying the Bayesian optimization, wherein all variations of the objective function are included in a linear subspace in a dimension less than the threshold value” (mental process and/or math concept), as drafted, is a process that, under its broadest reasonable interpretation, covers performance of the limitation in the mind but for the recitation of generic computer components. That is, other than reciting possible generic computer element for implementing the abstract idea, nothing in the claim element precludes the step from practically being performed in the mind. For example, but for the language about generic computer components, “performing”, in the limitation citied above encompasses performing evaluation to reduce dimensionality of certain function which is based on observation, evaluation, judgement, and/or opinion, that could be performed by human using paper / pen / calculator. If a claim limitation, under its broadest reasonable interpretation, covers performance of the limitation in the mind but for the recitation of generic computer components, then it falls within the “Mental Processes” grouping of abstract ideas. The claim does not include additional elements that shows integration into a practical application or adding something significantly more to the judicial exception. The claim is not patent eligible.
With regards to claim 6, the claim recites “wherein the dimension of the objective function is determined on the basis of the number of the plurality of convolutional layers” (math concept), as drafted, is a process that, under its broadest reasonable interpretation, covers performance of the limitation in the mind but for the recitation of generic computer components. That is, other than reciting possible generic computer element for implementing the abstract idea, nothing in the claim element precludes the step from practically being performed in the mind. For example, but for the language about generic computer components, “determined”, in the limitation citied above encompasses performing evaluation to reduce dimensionality of certain function which is based on observation, evaluation, judgement, and/or opinion, that could be performed by human using paper / pen / calculator. If a claim limitation, under its broadest reasonable interpretation, covers performance of the limitation in the mind but for the recitation of generic computer components, then it falls within the “Mental Processes” grouping of abstract ideas. The claim does not include additional elements that shows integration into a practical application or adding something significantly more to the judicial exception. The claim is not patent eligible.
With regards to claim 7, the claim recites “wherein the condition is defined as a formula below,
PNG
media_image1.png
457
935
media_image1.png
Greyscale
PNG
media_image2.png
112
926
media_image2.png
Greyscale
” (mental process and/or math concept), as drafted, is a process that, under its broadest reasonable interpretation, covers mathematical concept, which is an abstract idea. The claim is not patent eligible.
With regards to claim 8, the claim recites “wherein the projection is defined as a formula below,
PNG
media_image3.png
325
926
media_image3.png
Greyscale
” (mental process and/or math concept), as drafted, is a process that, under its broadest reasonable interpretation, covers mathematical concept, which is an abstract idea. The claim is not patent eligible.
Allowable Subject Matter Analysis
Claims 1-8 include allowable subject matter since when reading the claims in light of the specification, as per, MPEP §2111.01 or Toro Co. v. White Consolidated Industries Inc., 199F.3d 1295, 1301, 53 USPQ2d 1065, 1069, 1069 (Fed.Cir. 1999), none of the references of record alone or in combination disclose or suggest the combination of limitations specified in claims 1-8.
In interpreting the claims, in light of the specification filed on 27 December 2022, the Examiner finds the claimed invention to be patentably distinct from the prior arts of record.
Regarding independent claim 1, the primary reason for the allowance is the inclusion of the following elements, in combination with the other elements cited, which is not found in the prior art of record:
“… setting a condition of an objective function on the basis of combinations of pruning rates respectively applied to the plurality of convolutional layers, wherein the condition is that specifies the combination of pruning rates minimizing a value of the objective function minimizes a difference between filters of the plurality of convolutional layers and filters of the plurality of convolutional layers pruned by the combination of pruning rates minimizing the value of the objective function; and
determining the combination of pruning rates minimizing the value of the objective function as a combination of optimal pruning rates from the objective function on the basis of Bayesian optimization,
wherein the objective function is defined as a sum of weight change rates across all of the plurality of convolutional layers, the weight change rates resulting from pruning each of the plurality of convolutional layers, and
wherein the conditions of the objective function are that (i) a storage cost ratio between a filter before pruninq and a filter after pruninq is equal to or less than a first threshold value, and (ii) a computational cost ratio between the filter before pruning and the filter after pruning is equal to or less than a second threshold value”.
Regarding the dependent claims, which include all the limitations of the independent claims, are also allowed.
The followings are references close to the invention claimed:
Sundaresan et al., US-PGPUB NO.20220036194A1 [hereafter Sundaresan] shows neural network pruning with target pruning rate.
Choi et al., US-PGPUB NO.20190347554A1 [hereafter CHoi] shows convolutional neural network optimization with storage / computation ratio.
Wu et al., US-PATENT NO.11704536B1 [hereafter Wu shows optimal operator sequence generation for neural network processing.
Liebenwein, et al., “Provable filter pruning for efficient neural networks”, arXiv: 1911.07412v2 [cs.LG] 23 Mar 2020 [hereafter Liebenwein] shows Filter pruning for efficient neural networks.
Xu, et al., “ReForm: static and dynamic resource-aware DNN reconfiguration framework for mobile device”, DAC’19, 2019 [hereafter Xu] shows dynamic neural network pruning with resource consideration.
Conclusion
Any inquiry concerning this communication or earlier communications from the examiner should be directed to TSU-CHANG LEE whose telephone number is 571-272-3567. The fax number is 571-273-3567.
If attempts to reach the examiner by telephone are unsuccessful, the examiner’s supervisor, Omar Fernandez Rivas, can be reached 571-272-2589.
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://pair-direct.uspto.gov. Should you have questions on access to the Private PAIR system, contact the Electronic Business Center (EBC) at 866-217-9197 (toll-free). If you would like assistance from a USPTO Customer Service Representative or access to the automated information system, call 800-786-9199 (IN USA OR CANADA) or 571-272-1000.
/TSU-CHANG LEE/
Primary Examiner, Art Unit 2128