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and the fee set forth in 37 CFR 1.17(e) has been timely paid, the finality of the previous Office action has been withdrawn pursuant to 37 CFR 1.114.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 January 30, 2026 has been entered.
Claim Rejections - 35 USC § 112
The following is a quotation of the first paragraph of 35 U.S.C. 112(a):
(a) IN GENERAL.—The specification shall contain a written description of the invention, and of the manner and process of making and using it, in such full, clear, concise, and exact terms as to enable any person skilled in the art to which it pertains, or with which it is most nearly connected, to make and use the same, and shall set forth the best mode contemplated by the inventor or joint inventor of carrying out the invention.
The following is a quotation of the first paragraph of pre-AIA 35 U.S.C. 112:
The specification shall contain a written description of the invention, and of the manner and process of making and using it, in such full, clear, concise, and exact terms as to enable any person skilled in the art to which it pertains, or with which it is most nearly connected, to make and use the same, and shall set forth the best mode contemplated by the inventor of carrying out his invention.
Claims 20-24 are rejected under 35 U.S.C. 112(a) or 35 U.S.C. 112 (pre-AIA ), first paragraph, as failing to comply with the enablement requirement. The claim(s) contains subject matter which was not described in the specification in such a way as to enable one skilled in the art to which it pertains, or with which it is most nearly connected, to make and/or use the invention. The specification is silent with respect to “the second neural network include a plurality of candidate networks and an assistant network”. Paragraphs [082]-[088] of the instant specification discusses candidate networks but there is not mention of them being part of the second neural networks.
Claim Rejections - 35 USC § 103
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 (i.e., changing from AIA to pre-AIA ) 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.
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- 4, 7-10, 13-16, 17-22 and 25-28 are rejected under 35 U.S.C. 103 as being unpatentable over “Neural Architecture Search (NAS): basic principles and different approaches” by Sergios Karagiannakos (hereinafter Karagiannakos) in view of “In defense of weight-sharing for neural architecture search: an optimization perspective” by Khodak et al.
**Examiner note: When printing the NPL to PDF the website menu covered the text of the document, a clean printed version is appended to the end of the NPL”
With respect to claim 1:
Karagiannakos teaches:
A processor comprising:
One or more circuits to use one or more first neural networks to:
Obtain architectural information for a plurality of candidate network architectures to perform a task; (Page 1, discloses a plurality of network architectures specific for a task)
Use an assistant network to modify at least a portion of shared weights for the plurality of candidate network architectures based, at least in part, on input data for the task and the architectural information for at least a portion of the plurality of candidate network architectures; and (Page 1 and Page 9, disclose using a controller to modify weights for the architectures based on a specific task)
Select one of the candidate network architectures to generate a second neural network to perform the task based, at least in part, upon estimated performances of the task from training and evaluating the candidate network architectures using the at least a portion of shared weights modified by the assistant network. (Page 1, discloses selecting an optimal architecture for a specific task based on the performance of the ask and a validation test)
Karagiannakos doest not appear to explicitly disclose:
Shared weights.
Khodak teaches:
Shared weights. (Page 1, discloses weight-sharing, i.e. only one set of model parameters is trained for all candidate architectures)
It would have been obvious to one of ordinary skill in the art before the effective filing date of the present application to implement a method that utilized the teachings of Karagiannakos and the teachings of Khodak, both in the same field of invention of Neural Architecture Search. This would have provided the advantage of dramatically accelerating a Neural Architecture Search (Khodak, Page 1).
With respect to claim 2:
The combination of Karagiannakos and Khodak teaches:
The processor of claim 1, wherein the one or more first neural networks include an architecture search network to select one or more candidate architectures for the second neural network from an architecture search space that includes architectures for neural network types relevant to a set of tasks, including the task to be performed by the second neural network. (Page 1, discloses a neural architecture search that selects architectures for design of a neural network topology from a search space of possible network topologies)
With respect to claim 3:
The combination of Karagiannakos and Khodak teaches:
The processor of claim 2, wherein the architecture search network is configured to identify a set of candidate architectures from the architecture search space and update network parameters for the candidate architectures until an estimated performance of one of the candidate architectures for the task is determined to satisfy at least one criterion to select the candidate architecture for the second neural network. (Page 1, discloses the NAS choosing a list of possible architectures, ranking them based on their performance and determining the optimal architecture based on reaching a certain condition. Page 9 discloses recursively updating changing the model parameters until a condition is met)
With respect to claim 4:
The combination of Karagiannakos and Khodak teaches:
The processor of claim 2, wherein the architecture search network uses a HyperNet as the assistant network to determine one or more channel-wise weights based, at least in part, upon architecture topology data available during training. (Page 9, discloses a Hypernet that generates and modifies weights based on architectures)
Claims 7-10, 13-16, 17-22 and 25-28 are rejected in the same grounds as claims 1-4 above.
Claims 5, 11, 17, 23 and 29 are rejected under 35 U.S.C. 103 as being unpatentable over “Neural Architecture Search (NAS): basic principles and different approaches” by Sergios Karagiannakos (hereinafter Karagiannakos) in view of “In defense of weight-sharing for neural architecture search: an optimization perspective” by Khodak et al. and “Improving One-Shot NAS with Shrinking-and-Expanding Supernet” by Hu et al.
With respect to claim 5:
The combination of Karagiannakos and Khodak does not appear to disclose:
The processor of claim 4, wherein the one or more circuits are further to remove the HyperNet from the architecture search network using an annealing process.
Hu teaches:
The processor of claim 4, wherein the one or more circuits are further to remove the HyperNet from the architecture search network using an annealing process. (Hu, Page 4 Section 3.3, discloses search space shrinking by pruning weak operators)
It would have been obvious to one of ordinary skill in the art before the effective filing date of the present application to implement a method that utilized the teachings of Karagiannakos, Khodak and the teachings of Hu, both in the same field of invention of Neural Architecture Search. This would have provided the advantage of speeding up a neural architecture search (Hu, Abstract).
Claims 11, 17, 23 and 29 are rejected in the same grounds as claim 5 above.
Claim(s) 6, 12, 18, 24 and 30 are rejected under 35 U.S.C. 103 as being unpatentable over “Neural Architecture Search (NAS): basic principles and different approaches” by Sergios Karagiannakos (hereinafter Karagiannakos) in view of “In defense of weight-sharing for neural architecture search: an optimization perspective” by Khodak et al. and Pub. No. WO 2021183684 A1, hereinafter Curtin et al.
With respect to claim 6:
The combination of Karagiannakos and Khodak does not appear to disclose:
The processor of claim 1, wherein the task relates to three-dimensional medical image segmentation.
Curtin teaches:
The processor of claim 1, wherein the task relates to three-dimensional medical image segmentation. (Paragraph [16], discloses using a neural network for segmenting medical images)
It would have been obvious to one of ordinary skill in the art before the effective filing date of the present application to implement a method that utilized the teachings of Karagiannakos, Khodak and the teachings of Curtin, both in the same field of invention of Neural Networks. This would have provided the advantage of allowing the ability to detect abnormalities below the ability of a human interpreter’s perception (Paragraph [16]).
Claims 12, 18, 24 and 30 are rejected in the same grounds as claim 6 above.
Response to Arguments
Claim Rejections - 35 USC § 112
With respect to applicant’s arguments related to the 35 USC 112(a) rejections for claims 1-18 and 25-30, the amendments and arguments have overcome the previously set forth rejections.
With respect to applicant’s arguments related to the 35 USC 112(a) rejections for claims 20-24 that the amendments overcome the rejection. Examiner respectfully disagrees. It is not clear from the drawings (Figs. 2B, 3 and 5A) that a neural network includes an architecture search network.
With respect to applicant’s arguments related to the 35 USC 112(b) rejections, the amendments and arguments have overcome the previously set forth rejections.
Claim Rejections - 35 USC § 102
Applicant argues “The HyperNet in Karagiannakos generates new weight for each random chosen architecture. It does not modify shared weights”. The arguments are moot in view of the new grounds of rejection.
Applicant also argues “The HyperNet in Karagiannakos does not also use input data for the task as an additional basis to modify weights”.
Examiner respectfully disagrees. Kragiannakos (Page 1) discloses that the training of the candidate architectures is based on a specific task and these includes the input data to execute that task (validation test). It is well known that a validation test requires a validation data set with equates to the input data.
Conclusion
Any inquiry concerning this communication or earlier communications from the examiner should be directed to MARIELA D REYES whose telephone number is (571)270-1006. The examiner can normally be reached Monday-Friday, 7:30 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, David Wiley can be reached at (571) 272-4150. 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.
/Mariela Reyes/Supervisory Patent Examiner, Art Unit 2142
Prosecution Insights