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 FINAL and is in response to the amendment filed June 12th, 2026. Claims 1-24 are pending, of which claims 1-24 are currently rejected.
Response to Arguments
The amendment filed June 12th, 2026 has been entered. Claims 1-24 remain pending in the application. Applicant’s amendments to the Claims have overcome each and every claim objection as previously set forth in the Non-Final Office Action mailed April 6th, 2026.
Claim Objections
Applicant has amended the claims, and therefore the previous claim objections have been withdrawn.
Claim Rejections - 35 USC § 101
Arguments presented by Applicant are not persuasive.
Applicant alleges the recitation of “on a convolutional layer of a neural network” integrates the judicial exception into a practical application (Applicant Remarks: Pg. 18).
Examiner respectfully disagrees. At most this limitation generally links the claimed invention to the field of convolutional neural networks. This limitation does not integrate the judicial exception into a practical application.
Furthermore, Applicant alleges that because the claimed apparatus solves the technical problem of hardware utilization inefficiency, complex padding logic, or redundant data fetching from memory, thereby providing a technical improvement, that the this makes the claimed invention patent eligible (Applicant Remarks: Pg. 19).
Examiner respectfully disagrees. The improvement of the invention as claimed comes from the specific knitting and unknitting of matrices and the convolution that occurs with respect to his knitting and unknitting, which ultimately is math. It is important to keep in mind that an improvement in the abstract idea itself (e.g. a recited fundamental economic concept) is not an improvement in technology” (MPEP 2106.05(a)(II)). The 'inventive concept cannot be furnished by the unpatentable law or nature (or natural phenomenon or abstract idea) itself. MPEP 2106.05.I. See also MPEP 2106.05(a). The judicial exception alone cannot provide the improvement.
Applicant states that the functionality of the claimed invention is not well-understood, routine or conventional. However, what is being characterized as well-understood, routine and conventional in the Non-Final Office Action mailed April 6th, 2026 is the function of reading as recited in the claims. As discussed with respect to the rest of the claims, the invention as claimed comprises either additional elements that do not constitute a patent eligible invention and math.
See Claim Rejections - 35 USC § 101.
Prior Art Rejections
Applicant’s arguments regarding 103 rejections have been considered and are not persuasive.
Applicant alleges that Han does not teach a convolutional kernel being unknitted or split into a plurality of sub-kernels (Applicant Remarks: Pg. 22). Examiner respectfully disagrees. Han explicitly teaches the splitting of a convolutional kernel so that convolutions can occur corresponding to second matrices i.e., second block matrices as discussed in ¶ 0095 - ¶ 0096 and steps shown in Fig. 9 of Han. Therefore, Han does in fact teach a convolutional kernel being unknitted or split into a plurality of sub-kernels.
Applicant additionally alleges that Han merely teaches performing a general convolution, and does not teach s*s sub-kernels being applied to s*s second matrices (Applicant Remarks: Pg. 23). Examiner respectfully disagrees. Han teaches the convolutional kernel being applied to the corresponding second matrices (Han: ¶ 0034 - ¶ 0035), which is further elaborated upon in terms of the splitting of the convolutional kernel into sub kernels and applied to corresponding second block matrices i.e., second matrices (Applicant Remarks: Fig. 9, Figs. 6A and 6B, ¶ 0095 - ¶ 0096). Therefore, Han does in fact teach s*s sub-kernels being applied to s*s second matrices.
Applicant alleges that Kodavanji does not teach the s*s kernel elements being applied correspondingly on a one-to-one basis to the corresponding s*s second matrices (Applicant Remarks: Pg. 24). Examiner acknowledges what Applicant has alleged, however Han is relied upon to teach this aspect of the claimed invention (Han: Fig. 1A input feature from submatrix 11 being applied 1 to 1 to kernel 20; also discussed in ¶ 0034 - ¶ 0035), not Kodavanji.
Applicant additionally alleges that the stride size of M is not taught by Kodavanji, saying that Kodavanji does not teach such a feature (Applicant Remarks: Pg. 24). Examiner respectfully disagrees. Kodavanji teaches submatrices being split into MxM dimension matrices i.e., square matrices and stride/step size would be M, (Kodavanji: Fig. 2 shows M being 3 which is greater than 1; ¶ 0029; ¶ 0021 describes a step size of the submatrix being used for operations throughout main matrix).
There are new reasons for rejection as necessitated by amendments. See Claim Rejections – 35 USC § 102 and Claim Rejections – 35 USC § 103.
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 1-24 are rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more.
Regarding claim 1, at Step 1, the claim is directed to a statutory category of invention (machine).
At Step 2A, Prong 1, Examiner notes that the claim recites an abstract idea. Claim language recites partitioning matrices for a convolution operation with a stride greater than 1. Below are the limitations of claim 1 that recite an abstract idea under mathematical concepts:
to perform a convolution operation with a stride greater than 1
to unknit a first matrix into s*s second matrices or knit the s*s second matrices into the first matrix,
wherein the s is an integer greater than l and is the stride of the convolution operation,
the first matrix is split into a plurality of s*s subblocks,
and s*s pixels in each of the plurality of s*s sub blocks serve one-to-one as one pixel of the s*s second matrices;
unknits a convolution kernel used for performing the convolution operation with the stride of s on the first matrix into s*s sub-kernels according to the s*s pixels,
the s*s sub-kernels are applied one-to-one to the s*s second matrices, the convolution operation device uses any one of the s*s sub-kernels to perform a convolution operation with the stride of 1 on one corresponding second matrix among the s*s second matrices to generate a first operation result,
and accumulates the first operation result of each of the s*s second matrices as a second operation result of performing the convolution operation with the stride of s on the first matrix.
All limitations as indicated describe “mathematical concepts”.
At Step 2A Prong 2, these are the additional elements recited in claim 1:
A convolution apparatus
On a convolutional layer of a neural network
a data memory
a matrix unknit-knit device coupled to the data memory
a convolution operation device coupled to the data memory
These elements are generic computer components and do not integrate the judicial exception into a practical application of the exception. See MPEP 2106.05(f). These elements represent no more than mere instructions to apply the judicial exception on a computer and generally link the judicial exception to the field of convolutional layers in neural networks. Even when viewed in combination, these additional elements do not integrate the recited judicial exception into a practical application and the claim is directed to the judicial exception.
At Step 2B, there are no additional elements claimed that amount to significantly more than the recited judicial exception. The additional elements at best are the equivalent of merely adding the words “apply it” to the judicial exception. Mere instructions to apply an exception cannot provide an inventive concept.
Even when considered in combination, these additional elements represent mere instructions to apply an exception, which do not provide an inventive concept. The claim is not eligible.
Regarding claim 2, at Step 1, the claim is directed to a statutory category of invention (machine).
At Step 2A, Prong 1, Examiner notes that the claim recites an abstract idea. Below are the limitations of claim 2 that recite an abstract idea under mathematical concepts:
splits the first matrix into the plurality of s*s subblocks,
and collects pixels at a same position in the plurality of s*s subblocks as pixels of one of the s*s second matrices to unknit the first matrix into the s*s second matrices.
All limitations as indicated describe “mathematical concepts”.
At Step 2A Prong 2, these are the additional elements recited in claim 2:
wherein the matrix unknit-knit device reads the first matrix from the data memory (insignificant extra-solution activity)
These elements are generic computer components and do not integrate the judicial exception into a practical application of the exception. See MPEP 2106.05(f). These elements represent no more than mere instructions to apply the judicial exception on a computer. Even when viewed in combination, these additional elements do not integrate the recited judicial exception into a practical application and the claim is directed to the judicial exception.
At Step 2B, there are no additional elements claimed, whether alone or in combination, that amount to significantly more than the recited judicial exception. The additional elements at best are the equivalent of merely adding the words “apply it” to the judicial exception. Mere instructions to apply an exception cannot provide an inventive concept.
In regards to the insignificant extra-solution activity found in this limitation “wherein the matrix unknit-knit device reads the first matrix from the data memory”, this action describes mere data gathering that is recited at a high level of generality. Per MPEP 2106.05(d)(II), the courts have recognized the following computer functions as well‐understood, routine, and conventional functions when they are claimed in a merely generic manner (e.g., at a high level of generality) or as insignificant extra-solution activity: i. Receiving or transmitting data over a network, e.g., using the Internet to gather data, Symantec, 838 F.3d at 1321, 120 USPQ2d at 1362 (utilizing an intermediary computer to forward information); TLI Communications LLC v. AV Auto. LLC, 823 F.3d 607, 610, 118 USPQ2d 1744, 1745 (Fed. Cir. 2016) (using a telephone for image transmission); OIP Techs., Inc., v. Amazon.com, Inc., 788 F.3d 1359, 1363, 115 USPQ2d 1090, 1093 (Fed. Cir. 2015) (sending messages over a network); buySAFE, Inc. v. Google, Inc., 765 F.3d 1350, 1355, 112 USPQ2d 1093, 1096 (Fed. Cir. 2014) (computer receives and sends information over a network); iv. Storing and retrieving information in memory, Versata Dev. Group, Inc. v. SAP Am., Inc., 793 F.3d 1306, 1334, 115 USPQ2d 1681, 1701 (Fed. Cir. 2015); OIP Techs., 788 F.3d at 1363, 115 USPQ2d at 1092-93. This limitation therefore remains insignificant extra-solution activity even upon reconsideration. Thus, this limitation does not amount to significantly more.
Even when considered in combination, these additional elements represent mere instructions to apply an exception, which do not provide an inventive concept. The claim is not eligible.
Regarding claim 3, at Step 1, the claim is directed to a statutory category of invention (machine).
At Step 2A, Prong 1, Examiner notes that the claim recites an abstract idea. Below are the limitations of claim 3 that recite an abstract idea under mathematical concepts:
splits the first matrix into the plurality of s*s subblocks,
collects pixels at a same position in the s*s second matrices as pixels of one of the plurality of s*s sub blocks of the first matrix to knit the s*s second matrices into the first matrix.
All limitations as indicated describe “mathematical concepts”.
At Step 2A Prong 2, these are the additional elements recited in claim 3:
wherein the matrix unknit-knit device reads the first matrix from the data memory (insignificant extra-solution activity)
These elements are generic computer components and do not integrate the judicial exception into a practical application of the exception. See MPEP 2106.05(f). These elements represent no more than mere instructions to apply the judicial exception on a computer. Even when viewed in combination, these additional elements do not integrate the recited judicial exception into a practical application and the claim is directed to the judicial exception.
At Step 2B, there are no additional elements claimed, whether alone or in combination, that amount to significantly more than the recited judicial exception. The additional elements at best are the equivalent of merely adding the words “apply it” to the judicial exception. Mere instructions to apply an exception cannot provide an inventive concept.
In regards to the insignificant extra-solution activity found in this limitation “wherein the matrix unknit-knit device reads the first matrix from the data memory”, this action describes mere data gathering that is recited at a high level of generality. Per MPEP 2106.05(d)(II), the courts have recognized the following computer functions as well‐understood, routine, and conventional functions when they are claimed in a merely generic manner (e.g., at a high level of generality) or as insignificant extra-solution activity: i. Receiving or transmitting data over a network, e.g., using the Internet to gather data, Symantec, 838 F.3d at 1321, 120 USPQ2d at 1362 (utilizing an intermediary computer to forward information); TLI Communications LLC v. AV Auto. LLC, 823 F.3d 607, 610, 118 USPQ2d 1744, 1745 (Fed. Cir. 2016) (using a telephone for image transmission); OIP Techs., Inc., v. Amazon.com, Inc., 788 F.3d 1359, 1363, 115 USPQ2d 1090, 1093 (Fed. Cir. 2015) (sending messages over a network); buySAFE, Inc. v. Google, Inc., 765 F.3d 1350, 1355, 112 USPQ2d 1093, 1096 (Fed. Cir. 2014) (computer receives and sends information over a network); iv. Storing and retrieving information in memory, Versata Dev. Group, Inc. v. SAP Am., Inc., 793 F.3d 1306, 1334, 115 USPQ2d 1681, 1701 (Fed. Cir. 2015); OIP Techs., 788 F.3d at 1363, 115 USPQ2d at 1092-93. This limitation therefore remains insignificant extra-solution activity even upon reconsideration. Thus, this limitation does not amount to significantly more.
Even when considered in combination, these additional elements represent mere instructions to apply an exception, which do not provide an inventive concept. The claim is not eligible.
Regarding claim 4, at Step 1, the claim is directed to a statutory category of invention (machine).
At Step 2A, Prong 1, Examiner notes that the claim recites an abstract idea. Below are the limitations of claim 4 that recite an abstract idea under mathematical concepts:
wherein the stride s of the convolution operation is 2,
the first matrix is split into a plurality of 2*2 subblocks,
the 2*2 pixels in each of the plurality of 2*2 subblocks comprise an upper left pixel, an upper right pixel, a lower left pixel, and a lower right pixel,
the 2*2 second subblocks comprise a first unknitted matrix, a second unknitted matrix, a third unknitted matrix, and a fourth unknitted matrix,
the upper left pixel of the plurality of 2*2 subblocks serve as a pixel of the first unknitted matrix, the upper right pixel of the plurality of 2*2 subblocks serve as a pixel of the second unknitted matrix, the lower left pixel of the plurality of 2*2 subblocks serve as a pixel of the third unknitted matrix, and the lower right pixel of the plurality of 2*2 subblocks serve as a pixel of the fourth unknitted matrix.
All limitations as indicated describe “mathematical concepts”.
At Step 2A Prong 2, there are no additional elements beyond those recited in claim 1.
The claim is not eligible.
Regarding claim 5, at Step 1, the claim is directed to a statutory category of invention (machine).
At Step 2A, Prong 1, Examiner notes that the claim recites an abstract idea. Below are the limitations of claim 5 that recite an abstract idea under mathematical concepts:
wherein the stride s of the convolution operation is 3,
the first matrix is split into a plurality of 3 *3 sub blocks,
the 3 *3 pixels in each of the plurality of 3*3 subblocks comprise an upper left pixel, upper middle pixel, upper right pixel, middle left pixel, middle middle pixel, middle right pixel, lower left pixel, lower middle pixel, and lower right pixel,
the 3*3 second subblocks comprise a first unknitted matrix, a second unknitted matrix, a third unknitted matrix, a fourth unknitted matrix, a fifth unknitted matrix, a sixth unknitted matrix, a seventh unknitted matrix, an eighth unln1itted matrix, and a ninth unknitted matrix,
the upper left pixel of the plurality of 3*3 sub blocks serve as a pixel of the first unknitted matrix, the upper middle pixel of the plurality of 3*3 subblocks serve as a pixel of the second unknitted matrix, the upper right pixel of the plurality of 3*3 subb1ocks serve as a pixel of the third unknitted matrix, the middle left pixel of the plurality of 3*3 sub blocks serve as a pixel of 25 the fourth unknitted matrix, the middle middle pixel of the plurality of 3*3 subblocks serve as a pixel of the fifth unknitted matrix, the middle right pixel of the plurality of 3*3 subblocks serve as a pixel of the sixth unk11itted matrix, the lower left pixel of the plurality of 3*3 subblocks serve as a pixel of the seventh unknitted matrix, the lower middle pixel of the plurality of 3*3 sub blocks serve as a pixel of the eighth unknitted matrix, and the lower right pixel of the plurality of 3 *3 subblocks serve as a pixel of the ninth unknitted matrix.
All limitations as indicated describe “mathematical concepts”.
At Step 2A Prong 2, there are no additional elements beyond those recited in claim 1.
The claim is not eligible.
Regarding claim 6, at Step 1, the claim is directed to a statutory category of invention (machine).
At Step 2A, Prong 1, Examiner notes that the claim recites an abstract idea. Below are the limitations of claim 6 that recite an abstract idea under mathematical concepts:
wherein the stride s of the convolution operation is 2,
the convolution kernel is a 3 *3 matrix, the convolution kernel is unknitted into a first sub-kernel, a second sub-kernel, a third sub-kernel, and a fourth sub-kernel,
the first sub-kernel is a 2*2 matrix and comprises an upper left pixel, an upper right pixel, a lower left pixel, and a lower right pixel of the convolution kernel,
the second sub-kernel is a 2* 1 matrix and comprises an upper middle pixel and a lower middle pixel of the convolution kernel,
the third sub-kernel is 1*2 matrix and comprises a middle left pixel and a middle right pixel of the convolution kernel,
and the fourth sub-kernel is a 1 *1 matrix and comprises a middle middle pixel of the convolution kernel.
All limitations as indicated describe “mathematical concepts”.
At Step 2A Prong 2, there are no additional elements beyond those recited in claim 1.
The claim is not eligible.
Regarding claim 7, at Step 1, the claim is directed to a statutory category of invention (machine).
At Step 2A, Prong 1, Examiner notes that the claim recites an abstract idea. Below are the limitations of claim 7 that recite an abstract idea under mathematical concepts:
to unknit the first matrix into the s*s second matrices or knit the s*s second matrices into the first matrix.
All limitations as indicated describe “mathematical concepts”.
At Step 2A Prong 2, these are the additional elements recited in claim 7:
a temporary register configured to read the first matrix or the s*s second matrices from the data memory (insignificant extra-solution activity)
an execution unit coupled to the temporary register
These elements are generic computer components and do not integrate the judicial exception into a practical application of the exception. See MPEP 2106.05(f). These elements represent no more than mere instructions to apply the judicial exception on a computer. Even when viewed in combination, these additional elements do not integrate the recited judicial exception into a practical application and the claim is directed to the judicial exception.
At Step 2B, there are no additional elements claimed, whether alone or in combination, that amount to significantly more than the recited judicial exception. The additional elements at best are the equivalent of merely adding the words “apply it” to the judicial exception. Mere instructions to apply an exception cannot provide an inventive concept.
In regards to the insignificant extra-solution activity found in this limitation “a temporary register configured to read the first matrix or the s*s second matrices from the data memory”, this action describes mere data gathering that is recited at a high level of generality. Per MPEP 2106.05(d)(II), the courts have recognized the following computer functions as well‐understood, routine, and conventional functions when they are claimed in a merely generic manner (e.g., at a high level of generality) or as insignificant extra-solution activity: i. Receiving or transmitting data over a network, e.g., using the Internet to gather data, Symantec, 838 F.3d at 1321, 120 USPQ2d at 1362 (utilizing an intermediary computer to forward information); TLI Communications LLC v. AV Auto. LLC, 823 F.3d 607, 610, 118 USPQ2d 1744, 1745 (Fed. Cir. 2016) (using a telephone for image transmission); OIP Techs., Inc., v. Amazon.com, Inc., 788 F.3d 1359, 1363, 115 USPQ2d 1090, 1093 (Fed. Cir. 2015) (sending messages over a network); buySAFE, Inc. v. Google, Inc., 765 F.3d 1350, 1355, 112 USPQ2d 1093, 1096 (Fed. Cir. 2014) (computer receives and sends information over a network); iv. Storing and retrieving information in memory, Versata Dev. Group, Inc. v. SAP Am., Inc., 793 F.3d 1306, 1334, 115 USPQ2d 1681, 1701 (Fed. Cir. 2015); OIP Techs., 788 F.3d at 1363, 115 USPQ2d at 1092-93. This limitation therefore remains insignificant extra-solution activity even upon reconsideration. Thus, this limitation does not amount to significantly more.
Even when considered in combination, these additional elements represent mere instructions to apply an exception, which do not provide an inventive concept. The claim is not eligible.
Claims 8-14 recite the method practiced by the apparatus of claims 1-7 respectively and are therefore rejected under 35 U.S.C. 101 for the same reasons therein.
Regarding claim 15, at Step 1, the claim is directed to a statutory category of invention (machine).
At Step 2A, Prong 1, Examiner notes that the claim recites an abstract idea. Claim language recites partitioning of matrices for a convolution operation with a stride greater than 1. Below are the limitations of claim 15 that recite an abstract idea under mathematical concepts:
to perform a convolution operation with a stride greater than 1
to unknit the first matrix into the s*s second matrices or knit the s*s second matrices into the first matrix,
wherein the s is an integer greater than 1 and is the stride of the convolution operation,
the first matrix is split into a plurality of s*s sub blocks,
and s*s pixels in each of the plurality of s*s sub blocks serve one-to-one as one pixel of the s*s second matrices.
All limitations as indicated describe “mathematical concepts”.
At Step 2A Prong 2, these are the additional elements recited in claim 15:
A matrix unknit-knit device
On a convolutional layer of a neural network
a temporary register configured to read a first matrix or s*s second matrices from a data memory (insignificant extra-solution activity)
an execution unit coupled to the temporary register
These elements are generic computer components and do not integrate the judicial exception into a practical application of the exception. See MPEP 2106.05(f). These elements represent no more than mere instructions to apply the judicial exception on a computer and merely generally link the judicial exception to the field of convolutional layers in neural networks. Even when viewed in combination, these additional elements do not integrate the recited judicial exception into a practical application and the claim is directed to the judicial exception.
At Step 2B, there are no additional elements claimed, whether alone or in combination, that amount to significantly more than the recited judicial exception. The additional elements at best are the equivalent of merely adding the words “apply it” to the judicial exception. Mere instructions to apply an exception cannot provide an inventive concept.
In regards to the insignificant extra-solution activity found in this limitation “a temporary register configured to read the first matrix or the s*s second matrices from the data memory”, this action describes mere data gathering that is recited at a high level of generality. Per MPEP 2106.05(d)(II), the courts have recognized the following computer functions as well‐understood, routine, and conventional functions when they are claimed in a merely generic manner (e.g., at a high level of generality) or as insignificant extra-solution activity: i. Receiving or transmitting data over a network, e.g., using the Internet to gather data, Symantec, 838 F.3d at 1321, 120 USPQ2d at 1362 (utilizing an intermediary computer to forward information); TLI Communications LLC v. AV Auto. LLC, 823 F.3d 607, 610, 118 USPQ2d 1744, 1745 (Fed. Cir. 2016) (using a telephone for image transmission); OIP Techs., Inc., v. Amazon.com, Inc., 788 F.3d 1359, 1363, 115 USPQ2d 1090, 1093 (Fed. Cir. 2015) (sending messages over a network); buySAFE, Inc. v. Google, Inc., 765 F.3d 1350, 1355, 112 USPQ2d 1093, 1096 (Fed. Cir. 2014) (computer receives and sends information over a network); iv. Storing and retrieving information in memory, Versata Dev. Group, Inc. v. SAP Am., Inc., 793 F.3d 1306, 1334, 115 USPQ2d 1681, 1701 (Fed. Cir. 2015); OIP Techs., 788 F.3d at 1363, 115 USPQ2d at 1092-93. This limitation therefore remains insignificant extra-solution activity even upon reconsideration. Thus, this limitation does not amount to significantly more.
Even when considered in combination, these additional elements represent mere instructions to apply an exception, which do not provide an inventive concept. The claim is not eligible.
Regarding claim 16, at Step 1, the claim is directed to a statutory category of invention (machine).
At Step 2A, Prong 1, Examiner notes that the claim recites an abstract idea. Below are the limitations of claim 16 that recite an abstract idea under mathematical concepts:
splits the first matrix into the plurality of s*s subblocks,
collects pixels at a same position in the plurality of s*s subblocks as pixels of one of the s*s second matrices to unknit the first matrix into the s*s second matrices.
All limitations as indicated describe “mathematical concepts”.
At Step 2A Prong 2, these are the additional elements recited in claim 16:
wherein the execution unit reads the first matrix from the temporary register (insignificant extra-solution activity)
These elements are generic computer components and do not integrate the judicial exception into a practical application of the exception. See MPEP 2106.05(f). These elements represent no more than mere instructions to apply the judicial exception on a computer. Even when viewed in combination, these additional elements do not integrate the recited judicial exception into a practical application and the claim is directed to the judicial exception.
At Step 2B, there are no additional elements claimed, whether alone or in combination, that amount to significantly more than the recited judicial exception. The additional elements at best are the equivalent of merely adding the words “apply it” to the judicial exception. Mere instructions to apply an exception cannot provide an inventive concept.
In regards to the insignificant extra-solution activity found in this limitation “wherein the execution unit reads the first matrix from the temporary register”, this action describes mere data gathering that is recited at a high level of generality. Per MPEP 2106.05(d)(II), the courts have recognized the following computer functions as well‐understood, routine, and conventional functions when they are claimed in a merely generic manner (e.g., at a high level of generality) or as insignificant extra-solution activity: i. Receiving or transmitting data over a network, e.g., using the Internet to gather data, Symantec, 838 F.3d at 1321, 120 USPQ2d at 1362 (utilizing an intermediary computer to forward information); TLI Communications LLC v. AV Auto. LLC, 823 F.3d 607, 610, 118 USPQ2d 1744, 1745 (Fed. Cir. 2016) (using a telephone for image transmission); OIP Techs., Inc., v. Amazon.com, Inc., 788 F.3d 1359, 1363, 115 USPQ2d 1090, 1093 (Fed. Cir. 2015) (sending messages over a network); buySAFE, Inc. v. Google, Inc., 765 F.3d 1350, 1355, 112 USPQ2d 1093, 1096 (Fed. Cir. 2014) (computer receives and sends information over a network); iv. Storing and retrieving information in memory, Versata Dev. Group, Inc. v. SAP Am., Inc., 793 F.3d 1306, 1334, 115 USPQ2d 1681, 1701 (Fed. Cir. 2015); OIP Techs., 788 F.3d at 1363, 115 USPQ2d at 1092-93. This limitation therefore remains insignificant extra-solution activity even upon reconsideration. Thus, this limitation does not amount to significantly more.
Even when considered in combination, these additional elements represent mere instructions to apply an exception, which do not provide an inventive concept. The claim is not eligible.
Regarding claim 17, at Step 1, the claim is directed to a statutory category of invention (machine).
At Step 2A, Prong 1, Examiner notes that the claim recites an abstract idea. Below are the limitations of claim 17 that recite an abstract idea under mathematical concepts:
splits the first matrix into the plurality of s*s subblocks,
collects pixels at a same position in the s*s second matrices as pixels as of one of the plurality of s*s subblocks of the first matrix to knit the s*s second matrices into the first matrix.
All limitations as indicated describe “mathematical concepts”.
At Step 2A Prong 2, these are the additional elements recited in claim 17:
wherein the execution unit reads the s*s second matrices from the temporary register (insignificant extra-solution activity)
These elements are generic computer components and do not integrate the judicial exception into a practical application of the exception. See MPEP 2106.05(f). These elements represent no more than mere instructions to apply the judicial exception on a computer. Even when viewed in combination, these additional elements do not integrate the recited judicial exception into a practical application and the claim is directed to the judicial exception.
At Step 2B, there are no additional elements claimed, whether alone or in combination, that amount to significantly more than the recited judicial exception. The additional elements at best are the equivalent of merely adding the words “apply it” to the judicial exception. Mere instructions to apply an exception cannot provide an inventive concept.
In regards to the insignificant extra-solution activity found in this limitation “wherein the execution unit reads the s*s second matrices from the temporary register”, this action describes mere data gathering that is recited at a high level of generality. Per MPEP 2106.05(d)(II), the courts have recognized the following computer functions as well‐understood, routine, and conventional functions when they are claimed in a merely generic manner (e.g., at a high level of generality) or as insignificant extra-solution activity: i. Receiving or transmitting data over a network, e.g., using the Internet to gather data, Symantec, 838 F.3d at 1321, 120 USPQ2d at 1362 (utilizing an intermediary computer to forward information); TLI Communications LLC v. AV Auto. LLC, 823 F.3d 607, 610, 118 USPQ2d 1744, 1745 (Fed. Cir. 2016) (using a telephone for image transmission); OIP Techs., Inc., v. Amazon.com, Inc., 788 F.3d 1359, 1363, 115 USPQ2d 1090, 1093 (Fed. Cir. 2015) (sending messages over a network); buySAFE, Inc. v. Google, Inc., 765 F.3d 1350, 1355, 112 USPQ2d 1093, 1096 (Fed. Cir. 2014) (computer receives and sends information over a network); iv. Storing and retrieving information in memory, Versata Dev. Group, Inc. v. SAP Am., Inc., 793 F.3d 1306, 1334, 115 USPQ2d 1681, 1701 (Fed. Cir. 2015); OIP Techs., 788 F.3d at 1363, 115 USPQ2d at 1092-93. This limitation therefore remains insignificant extra-solution activity even upon reconsideration. Thus, this limitation does not amount to significantly more.
Even when considered in combination, these additional elements represent mere instructions to apply an exception, which do not provide an inventive concept. The claim is not eligible.
Regarding claim 18, at Step 1, the claim is directed to a statutory category of invention (machine).
At Step 2A, Prong 1, Examiner notes that the claim recites an abstract idea. Below are the limitations of claim 18 that recite an abstract idea under mathematical concepts:
wherein the stride s is 2,
the first matrix is split into a plurality of 2*2 subblocks,
the 2*2 pixels in each of the plurality of 2*2 subblocks comprise an upper left pixel, an upper right pixel, a lower left pixel, and a lower right pixel,
the 2*2 second matrices comprise a first unknitted matrix, a second unknitted matrix, a third unknitted matrix, and a fourth unknitted matrix,
the upper left pixel of the plurality of 2*2 subblocks serve as a pixel of the first unknitted matrix, the upper right pixel of the plurality of 2*2 subblocks serve as a pixel of the second unknitted matrix, the lower left pixel of the plurality of 20 2*2 subblocks serve as a pixel of the third unknitted matrix, and the lower right pixel of the plurality of 2*2 subblocks serve as a pixel of the fourth unknitted matrix.
All limitations as indicated describe “mathematical concepts”.
At Step 2A Prong 2, there are no additional elements beyond those recited in claim 15.
The claim is not eligible.
Regarding claim 19, at Step 1, the claim is directed to a statutory category of invention (machine).
At Step 2A, Prong 1, Examiner notes that the claim recites an abstract idea. Below are the limitations of claim 19 that recite an abstract idea under mathematical concepts:
wherein the stride s is 3,
the first matrix is split into a plurality of 3*3 subblocks,
the 3*3 pixels in each of the plurality of 3*3 subblocks comprise an upper left pixel, upper middle pixel, upper right pixel, middle left pixel, middle middle pixel, middle right pixel, lower left pixel, lower middle pixel, and lower right pixel,
the 3*3 second matrices comprise a first unknitted matrix, a second unknitted matrix, a third unknitted matrix, a fourth unknitted matrix, a fifth unknitted matrix, a sixth unknitted matrix, a seventh unknitted matrix, an eighth unknitted matrix, and a ninth unknitted matrix,
the upper left pixel of the plurality of 3 *3 sub blocks serve as a pixel of the first unknitted matrix, the upper middle pixel of the plurality of 3 *3 subblocks serve as a pixel of the second unknitted matrix, the upper right pixel of the plurality of 3*3 sub blocks serve as a pixel of the third unknitted matrix, the middle left pixel of the plurality of 3 *3 subblocks serve as a pixel of the fourth unknitted matrix, the middle middle pixel of the plurality of 3*3 subblocks serve as a pixel of the fifth unknitted matrix, the middle right pixel of the plurality of 3 *3 sub blocks serve as a pixel of the sixth unknitted matrix, the lower left pixel of the plurality of 3*3 sub blocks serve as a pixel of the seventh unknitted matrix, the lower middle pixel of the plurality of 3*3 sub blocks serve as a pixel of the eighth unknitted matrix, and the lower right pixel of the plurality of3*3 subblocks serve as a pixel of the ninth unknitted matrix.
All limitations as indicated describe “mathematical concepts”.
At Step 2A Prong 2, there are no additional elements beyond those recited in claim 15.
The claim is not eligible.
Claims 20-24 recite the method practiced by the apparatus recited in claims 15-19 respectively and are therefore rejected under 35 U.S.C. 101 for the same reasons therein.
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 1-2, 5, 7-9, 12, 14-16, 19-21, and 24 are rejected under 35 U.S.C. 103 as being unpatentable over Han (US 2021/0011970 A1) (hereinafter “Han”), further in view of Kodavanji et al. US 2021/0334335 (hereinafter “Kodavanji”).
Regarding claim 1, Han teaches:
A convolution apparatus configured to perform a convolution operation with a stride greater than 1 on a convolutional layer of a neural network, the convolution apparatus comprising:
a data memory (Han: Fig. 2 Element 70 as data memory; ¶ 0040 memory 70 for storing data for operation of neural network accelerator);
a matrix unknit-knit device coupled to the data memory (Han: Fig. 2 Element 80 neural network accelerator as matrix knit-unknit device) and configured to unknit a first matrix stored in the data memory into s*s second matrices or knit the s*s second matrices stored in the data memory into the first matrix (Han: ¶ 0011 first matrix of M×K dimensions being split into multiple sub-matrices i.e., second matrices);
and a convolution operation device coupled to the data memory (Han: Fig. 2 element 86), wherein the convolution operation device unknits a convolution kernel used for performing the convolution operation (Han: ¶ 0012 convolution device unknits the matrices into submatrices)
the s*s sub-kernels are applied one-to-one to the s*s second matrices (Han: Fig. 1A input feature from submatrix 11 being applied 1 to 1 to kernel 20; also discussed in ¶ 0034 - ¶ 0035), the convolution operation device uses any one of the s*s sub-kernels to perform a convolution operation with a stride of 1 on one corresponding second matrix among the s*s second matrices to generate a first operation result (Han: Fig. 1A element 30 is convolution result i.e., first operation result),
and the convolution operation device accumulates the first operation result of each of the s*s second matrices as a second operation result of performing the convolution operation with the stride of s on the first matrix (Han: ¶ 0059 after convolution, accumulation of convolution operations occurs in ACC accumulation register).
Han does not explicitly teach:
wherein the s is an integer greater than l and is the stride of the convolution operation, the first matrix is split into a plurality of s*s subblocks, and s*s pixels in each of the plurality of s*s sub blocks serve one-to-one as one pixel of the s*s second matrices;
for performing the convolution operation with a stride of s on the first matrix into s*s sub-kernels according to the s*s pixels.
However, Kodavanji teaches:
wherein the s is an integer greater than l and is the stride of the convolution operation, the first matrix is split into a plurality of s*s subblocks, and s*s pixels in each of the plurality of s*s sub blocks serve one-to-one as one pixel of the s*s second matrices (Kodavanji: submatrices being split into MxM dimension matrices i.e., square matrices and stride/step size would be M, Fig. 3 shows M being 3 which is greater than 1);
for performing the convolution operation with a stride of s on the first matrix into s*s sub-kernels according to the s*s pixels (Kodavanji: submatrices being split into MxM dimension matrices i.e., square matrices and stride/step size would be M).
It would be obvious before the effective filing date of the claimed invention to combine the dimension based stride as taught by Kodavanji with the unknitting and convolution operation as taught by Han as both references are directed towards matrix convolution operations. One with ordinary skill in the art would be motivated to combine the references because doing so would optimize allocation of inputs to memory spaces (Kodavanji: ¶ 0058).
Regarding claim 2, Han teaches:
The convolution apparatus according to claim 1, wherein the matrix unknit-knit device reads the first matrix from the data memory (Han: Fig. 2 Element 70; ¶ 0040 memory 70 for storing data for operation of neural network accelerator, first matrix would have to be accessed from data memory for operations).
While Han does teach splitting the matrix into second matrices, Han does not explicitly teach:
the matrix unknit-knit device splits the first matrix into the plurality of s*s subblocks, and the matrix unknit-knit device collects pixels at a same position in the plurality of s*s subblocks as pixels of one of the s*s second matrices to unknit the first matrix into the s*s second matrices.
However, Kodavanji teaches:
the matrix unknit-knit device splits the first matrix into the plurality of s*s subblocks, and the matrix unknit-knit device collects pixels at a same position in the plurality of s*s subblocks as pixels of one of the s*s second matrices to unknit the first matrix into the s*s second matrices (Kodavanji: ¶ 0013 submatrices being split into MxM dimension matrices and stride/step size would be M, Fig. 3 shows M being 3 which is greater than 1).
The motivation to combine with respect to claim 1 applies equally to claim 2.
Regarding claim 5, while Han teaches splitting of matrix into various sub-matrices for convolution operations, Han does not explicitly teach the first matrix being split into 9 sub-matrices.
However, Kodavanji teaches:
The convolution apparatus according to claim 1, wherein the stride s of the convolution operation is 3 (Kodavanji: ¶ 0013 submatrices being split into MxM dimension matrices and stride/step size would be M, Fig. 3 shows M being 3 which is greater than 1; exemplary partitioning is shown in Fig. 2 having convolutions occurring with stride M=3),
the first matrix is split into a plurality of 3 *3 sub blocks (Kodavanji: Fig. 2),
the 3 *3 pixels in each of the plurality of 3*3 subblocks comprise an upper left pixel, upper middle pixel, upper right pixel, middle left pixel, middle middle pixel, middle right pixel, lower left pixel, lower middle pixel, and lower right pixel (Kodavanji: Fig. 2 each of the subblocks comprise nine pixels),
the 3*3 second matrices comprise a first unknitted matrix, a second unknitted matrix, a third unknitted matrix, a fourth unknitted matrix, a fifth unknitted matrix, a sixth unknitted matrix, a seventh unknitted matrix, an eighth unknitted matrix, and a ninth unknitted matrix (Kodavanji: Fig. 2 first through ninth unknitted matrices are present after partitioning),
the upper left pixel of the plurality of 3*3 sub blocks serve as a pixel of the first unknitted matrix, the upper middle pixel of the plurality of 3*3 subblocks serve as a pixel of the second unknitted matrix, the upper right pixel of the plurality of 3*3 subblocks serve as a pixel of the third unknitted matrix, the middle left pixel of the plurality of 3*3 sub blocks serve as a pixel of the fourth unknitted matrix, the middle middle pixel of the plurality of 3*3 subblocks serve as a pixel of the fifth unknitted matrix, the middle right pixel of the plurality of 3*3 subblocks serve as a pixel of the sixth unk11itted matrix, the lower left pixel of the plurality of 3*3 subblocks serve as a pixel of the seventh unknitted matrix, the lower middle pixel of the plurality of 3*3 sub blocks serve as a pixel of the eighth unknitted matrix, and the lower right pixel of the plurality of 3*3 subblocks serve as a pixel of the ninth unknitted matrix (Kodavanji: Fig. 2 shows corresponding subblocks and pixels as partitioned).
The motivation to combine with respect to claim 1 applies equally to claim 5.
Regarding claim 7, while Han teaches the execution unit for unknitting the first matrix into second matrices (Han: ¶ 0012 convolution device unknits matrices into submatrices; Fig. 2 element 86), Han does not teach a temporary register coupled to the execution unit.
However, Kodavanji teaches:
a temporary register configured to read the first matrix or the s*s second matrices from the data memory (Kodavanji: ¶ 0024 inputs including from first matrix provided to I/O registers before being partitioned); and
an execution unit coupled to the temporary register (Kodavanji: ¶ 0024 I/O registers being coupled to execution unit for partitioning).
Claims 8-9, 12 and 14 recite the method practiced by the apparatus of claims 1-2, 5 and 7 respectively, and are therefore rejected for the same reasons therein.
Regarding claim 15, Han teaches:
A matrix unknit-knit device configured to perform a convolution operation with a stride greater than 1 on a convolutional layer of a neural network, wherein the matrix unknit-knit device comprises:
to read a first matrix or s*s second matrices from a data memory (Han: Fig. 2 Element 70 as data memory; ¶ 0040 memory 70 for storing data for operation of neural network accelerator); and
an execution unit (Han: Fig. 2 Element 80 neural network accelerator as matrix knit-unknit device) configured to unknit the first matrix stored in the temporary register into the s*s second matrices or knit the s*s second matrices into the first matrix (Han: ¶ 0011 first matrix of M×K dimensions being split into multiple sub-matrices i.e., second matrices),
the first matrix is split into a plurality of s*s sub blocks, and s*s pixels in each of the plurality of s*s sub blocks serve one-to-one as one pixel of the s*s second matrices (Han: ¶ 0011 first matrix of M×K dimensions being split into multiple sub-matrices i.e., second matrices; Fig. 1A input feature from submatrix 11 being applied 1 to 1 to kernel 20; also discussed in ¶ 0034 - ¶ 0035).
Han does not explicitly teach a dimension-based stride or a temporary register that matrices are read from.
However, Kodavanji teaches a dimension-based stride (Kodavanji: ¶ 0013 submatrices being split into MxM dimension matrices and stride/step size would be M, Fig. 3 shows M being 3 which is greater than 1; exemplary partitioning is shown in Fig. 2 having convolutions occurring with stride M=3) and a temporary register that matrices are read from (Kodavanji: ¶ 0024 inputs including from first matrix provided to I/O registers before being partitioned).
The motivation to combine with respect to claim 1 applies equally to claim 15.
Regarding claim 16, Han in view of Kodavanji further teaches:
The matrix unknit-knit device according to claim 15, wherein the execution unit reads the first matrix from the temporary register (Kodavanji: ¶ 0024 inputs including from first matrix provided to I/O registers before being partitioned),
the execution unit splits the first matrix into the plurality of s*s subblocks, and the execution unit collects pixels at a same position in the plurality of s*s subblocks as pixels of one of the s*s second matrices to unknit the first matrix into the s*s second matrices (Kodavanji: ¶ 0013 submatrices being split into MxM dimension matrices and stride/step size would be M, Fig. 3 shows M being 3 which is greater than 1).
The motivation to combine with respect to claim 1 applies equally to claim 16.
Regarding claim 19, while Han teaches splitting of matrix into various sub-matrices for convolution operations, Han does not explicitly teach the first matrix being split into 9 sub-matrices.
However, Kodavanji teaches:
The matrix unknit-knit device according to claim 15, wherein the stride s is 3 (Kodavanji: ¶ 0013 submatrices being split into MxM dimension matrices and stride/step size would be M, Fig. 3 shows M being 3 which is greater than 1; exemplary partitioning is shown in Fig. 2 having convolutions occurring with stride M=3),
the first matrix is split into a plurality of 3 *3 sub blocks (Kodavanji: Fig. 2),
the 3 *3 pixels in each of the plurality of 3*3 subblocks comprise an upper left pixel, upper middle pixel, upper right pixel, middle left pixel, middle middle pixel, middle right pixel, lower left pixel, lower middle pixel, and lower right pixel (Kodavanji: Fig. 2 each of the subblocks comprise nine pixels),
the 3*3 second matrices comprise a first unknitted matrix, a second unknitted matrix, a third unknitted matrix, a fourth unknitted matrix, a fifth unknitted matrix, a sixth unknitted matrix, a seventh unknitted matrix, an eighth unknitted matrix, and a ninth unknitted matrix (Kodavanji: Fig. 2 first through ninth unknitted matrices are present after partitioning),
the upper left pixel of the plurality of 3*3 sub blocks serve as a pixel of the first unknitted matrix, the upper middle pixel of the plurality of 3*3 subblocks serve as a pixel of the second unknitted matrix, the upper right pixel of the plurality of 3*3 subblocks serve as a pixel of the third unknitted matrix, the middle left pixel of the plurality of 3*3 sub blocks serve as a pixel of the fourth unknitted matrix, the middle middle pixel of the plurality of 3*3 subblocks serve as a pixel of the fifth unknitted matrix, the middle right pixel of the plurality of 3*3 subblocks serve as a pixel of the sixth unk11itted matrix, the lower left pixel of the plurality of 3*3 subblocks serve as a pixel of the seventh unknitted matrix, the lower middle pixel of the plurality of 3*3 sub blocks serve as a pixel of the eighth unknitted matrix, and the lower right pixel of the plurality of 3*3 subblocks serve as a pixel of the ninth unknitted matrix (Kodavanji: Fig. 2 shows corresponding subblocks and pixels as partitioned).
The motivation to combine with respect to claim 1 applies equally to claim 19.
Claims 20-21 and 24 recite the method practiced by the apparatus of claims 15-16 and 19 respectively and are therefore rejected for the same reasons therein.
Claims 3, 10, 17 and 22 are rejected under 35 U.S.C. 103 as being unpatentable over Han in view of Kodavanji, further in view of Song et al. (US 2021/0117791 A1) (hereinafter “Song”).
Regarding claim 3, Han teaches:
The convolution apparatus according to claim 1, wherein the matrix unknit-knit device reads the first matrix from the data memory (Han: Fig. 2 Element 70; ¶ 0040 memory 70 for storing data for operation of neural network accelerator, first matrix would have to be accessed from data memory for operations).
While Han does teach splitting the matrix into second matrices, Han does not explicitly teach:
the matrix unknit-knit device splits the first matrix into the plurality of s*s subblocks, and the matrix unknit-knit device collects pixels at a same position in the plurality of s*s subblocks as pixels of one of the s*s second matrices to knit the first matrix into the s*s second matrices.
However, Kodavanji teaches:
the matrix unknit-knit device splits the first matrix into the plurality of s*s subblocks, and the matrix unknit-knit device collects pixels at a same position in the plurality of s*s subblocks as pixels of one of the s*s second matrices (Kodavanji: ¶ 0013 submatrices being split into MxM dimension matrices and stride/step size would be M, Fig. 3 shows M being 3 which is greater than 1).
The motivation to combine with respect to claim 1 applies equally to claim 3.
Han in view of Kodavanji does not teach:
to knit the s*s second matrices into the first matrix.
However, Song teaches deconvolution and re-merging of partitioned matrices after convolution operations have taken place (Song: Fig. 3 and Fig. 5 Merge operation after convolution operation, and Fig. 8 shows more on the Merge process).
It would be obvious before the effective filing date of the claimed invention to combine the deconvolution and merging of Song with the convolution apparatus and splitting of matrices as taught by Han in view of Kodavanji as all teachings are directed towards matrix convolution operations. One with ordinary skill in the art would be motivated to combine the teachings because doing so increases processing efficiency, increases processing speed, and thus reduces operation costs and increases the efficiency of the computing (Song: ¶ 0075).
Claim 10 recites the method practiced by the apparatus of claim 3 and is therefore rejected for the same reasons therein.
Regarding claim 17, Han in view of Kodavanji further teaches:
The matrix unknit-knit device according to claim 15, wherein the execution unit reads the first matrix from the temporary register (Kodavanji: ¶ 0024 inputs including from first matrix provided to I/O registers before being partitioned),
the execution unit splits the first matrix into the plurality of s*s subblocks, and the execution unit collects pixels at a same position in the plurality of s*s subblocks as pixels of one of the s*s second matrices (Kodavanji: ¶ 0013 submatrices being split into MxM dimension matrices and stride/step size would be M, Fig. 3 shows M being 3 which is greater than 1).
The motivation to combine with respect to claim 1 applies equally to claim 17.
Han in view of Kodavanji does not teach:
to knit the s*s second matrices into the first matrix.
However, Song teaches deconvolution and re-merging of partitioned matrices after convolution operations have taken place (Song: Fig. 3 and Fig. 5 Merge operation after convolution operation, and Fig. 8 shows more on the Merge process).
The motivation to combine with respect to claim 3 applies equally to claim 17.
Claim 22 recites the method practiced by the apparatus of claim 17 and is therefore rejected for the same reasons therein.
Claims 4, 6, 11, 13, 18, and 23 are rejected under 35 U.S.C. 103 as being unpatentable over Han in view of Kodavanji, further in view of Liu et al. (US 2023/0075264) (hereinafter “Liu”).
Regarding claim 4, while Han in view of Kodavanji teaches the convolution apparatus of claim 1 as well as a dimension-dependent stride (Kodavanji: ¶ 0013 submatrices being split into MxM dimension matrices and stride/step size would be M, Fig. 3 shows M being 3 which is greater than 1; exemplary partitioning is shown in Fig. 2 having convolutions occurring with stride M=3), Han in view of Kodavanji does not explicitly teach the specific 2*2 partitioning as claimed in claim 4.
However, Liu teaches:
the first matrix is split into a plurality of 2*2 subblocks (Liu: Fig. 2A first matrix is split into a plurality of 2*2 subblocks),
the 2*2 pixels in each of the plurality of 2*2 subblocks comprise an upper left pixel, an upper right pixel, a lower left pixel, and a lower right pixel (Liu: Fig. 2A each of the subblocks comprise the four pixels as claimed),
the 2*2 second matrices comprise a first unknitted matrix, a second unknitted matrix, a third unknitted matrix, and a fourth unknitted matrix (Liu: Fig. 2A there are four unknitted matrices after partitioning),
the upper left pixel of the plurality of 2*2 sub blocks serve as a pixel of the first unknitted matrix (Liu: Fig. 2A first unknitted matrix 0206 has upper left pixel 1),
the upper right pixel of the plurality of 2*2 subblocks serve as a pixel of the second unknitted matrix (Liu: Fig. 2A second unknitted matrix 0205 has upper right pixel 4)
the lower left pixel of the plurality of 2*2 subblocks serve as a pixel of the third unknitted matrix (Liu: Fig. 2A third unknitted matrix 0204 has lower left pixel 13),
and the lower right pixel of the plurality of 2*2 subblocks serve as a pixel of the fourth unknitted matrix (Liu: Fig. 2A fourth unknitted matrix 0203 has lower right pixel 16).
It would be obvious before the effective filing date of the claimed invention to combine the 2*2 unknitting as taught by Liu with the convolution apparatus of Han in view of Kodavanji as all teachings are directed towards matrix convolution. One with ordinary skill in the art would be motivated to combine the teachings because this would reduce memory consumption and computation (Liu: ¶ 0081).
Regarding claim 6, while Han in Kodavanji teaches the convolution apparatus of claim 1, Han in view of Kodavanji do not teach the specific unknitting as recited in claim 6.
However, Liu teaches:
wherein the stride s of the convolution operation is 2 (Liu: Fig. 4A shows unknitting and is described in ¶ 0048 having a stride of 2),
the convolution kernel is a 3 *3 matrix, (Liu: Fig. 4A element 0401 is a 3 by 3 kernel)
the convolution kernel is unknitted into a first sub-kernel, a second sub-kernel, a third sub-kernel, and a fourth sub-kernel (Liu: Fig. 4A the first through fourth sub-kernels 0403-0406 respectively),
the first sub-kernel is a 2*2 matrix and comprises an upper left pixel, an upper right pixel, a lower left pixel, and a lower right pixel of the convolution kernel (Liu: Fig. 4A 0403),
the second sub-kernel is a 2* 1 matrix and comprises an upper middle pixel and a lower middle pixel of the convolution kernel (Liu: Fig. 4A 0404),
the third sub-kernel is a 1 *2 matrix and comprises a middle left pixel and a middle right pixel of the convolution kernel (Liu: Fig. 4A 0405),
and the fourth sub-kernel is a 1 *1 matrix and comprises a middle middle pixel of the convolution kernel (Liu: Fig. 4A 0406).
The motivation to combine with respect to claim 4 applies equally to claim 6.
Claims 11 and 13 recite the methods practiced by the apparatus of claims 4 and 6 and are therefore rejected for the same reasons therein.
Regarding claim 18, while Han in view of Kodavanji teaches the matrix unknit-knit device of claim 15 as well as a dimension-dependent stride (Kodavanji: ¶ 0013 submatrices being split into MxM dimension matrices and stride/step size would be M, Fig. 3 shows M being 3 which is greater than 1; exemplary partitioning is shown in Fig. 2 having convolutions occurring with stride M=3), Han in view of Kodavanji does not explicitly teach the specific 2*2 partitioning as claimed in claim 18.
However, Liu teaches:
the first matrix is split into a plurality of 2*2 subblocks (Liu: Fig. 2A first matrix is split into a plurality of 2*2 subblocks),
the 2*2 pixels in each of the plurality of 2*2 subblocks comprise an upper left pixel, an upper right pixel, a lower left pixel, and a lower right pixel (Liu: Fig. 2A each of the subblocks comprise the four pixels as claimed),
the 2*2 second matrices comprise a first unknitted matrix, a second unknitted matrix, a third unknitted matrix, and a fourth unknitted matrix (Liu: Fig. 2A there are four unknitted matrices after partitioning),
the upper left pixel of the plurality of 2*2 sub blocks serve as a pixel of the first unknitted matrix (Liu: Fig. 2A first unknitted matrix 0206 has upper left pixel 1),
the upper right pixel of the plurality of 2*2 subblocks serve as a pixel of the second unknitted matrix (Liu: Fig. 2A second unknitted matrix 0205 has upper right pixel 4)
the lower left pixel of the plurality of 2*2 subblocks serve as a pixel of the third unknitted matrix (Liu: Fig. 2A third unknitted matrix 0204 has lower left pixel 13),
and the lower right pixel of the plurality of 2*2 subblocks serve as a pixel of the fourth unknitted matrix (Liu: Fig. 2A fourth unknitted matrix 0203 has lower right pixel 16).
The motivation to combine with respect to claim 4 applies equally to claim 18.
Claim 23 recites the method practiced by the apparatus of claim 18 and is therefore rejected for the same reason therein.
Conclusion
THIS ACTION IS MADE FINAL. Applicant is reminded of the extension of time policy as set forth in 37 CFR 1.136(a).
A shortened statutory period for reply to this final action is set to expire THREE MONTHS from the mailing date of this action. In the event a first reply is filed within TWO MONTHS of the mailing date of this final action and the advisory action is not mailed until after the end of the THREE-MONTH shortened statutory period, then the shortened statutory period will expire on the date the advisory action is mailed, and any nonprovisional extension fee (37 CFR 1.17(a)) pursuant to 37 CFR 1.136(a) will be calculated from the mailing date of the advisory action. In no event, however, will the statutory period for reply expire later than SIX MONTHS from the mailing date of this final action.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to MARIA DE JESUS RIVERA whose telephone number is (571)272-2793. The examiner can normally be reached Monday-Friday 7:30AM-5PM.
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, James Trujillo can be reached at (571) 272-3677. 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.
/M.D.R./Examiner, Art Unit 2151
/EMILY E LAROCQUE/ Primary Examiner, Art Unit 2182
Prosecution Insights