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 .
DETAILED ACTION
This action is responsive to the following communication: Non-Provisional Application filed Dec. 13, 2023.
Claims 1-20 are pending in the case. Claims 1, 17 and 20 are independent claims.
Claim Objections
Claims 2-16 and 18-19 objected to because of their dependency from rejected independent claims 1, 17. Claims 2-16 and 18-19 would be allowable if written in independent form.
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, 17 and 20 are rejected under 35 U.S.C. 103 as being unpatentable over Asad et al. (hereinafter Asad) U.S. Patent Publication No. 2022/0044098 in view of Hofer et al. (hereinafter Hofer) U.S. Patent Publication No. 2023/0177325 and in further view of (hereinafter PyTorch) “How to split tensors with overlap and then reconstruct the original tensor? - vision - PyTorch Forums” (hereinafter PyTorch) Retrieved from https://discuss.pytorch.org/t/how-to-split-tensors-with-overlap-and-then-reconstruct-the-original-tensor/70261.
With respect to independent claim 1, Asad teaches a method of implementing in hardware a dynamic neural network for operation on an input tensor (see e.g., Para [5][29][51]-“There is provided a method of implementing in hardware a recurrent neural network (RNN) for operation on a sequence of inputs, each step of the recurrent neural network being for operation on a different input of the sequence”” represent the predetermined plurality of inputs as one or more tensors of first dimensions and the sets of weights as one or more tensors of second dimensions, the first and second dimensions being selected to enable each group of non-causal operations to be performed in parallel as a convolution operation at the processing elements”), the method comprising:
receiving a representation of the dynamic neural network (see e.g., Para [111] [[112][123] - “At 801, a representation of the RNN 338 to be implemented in hardware is received at the transformation unit. The RNN representation may be represented in any suitable manner—such as a mathematical representation, or any other representation of the RNN on which the transformation unit is configured to operate.”);
transforming the representation of the dynamic neural network into a static network adapted to operate on a fixed size input (see e.g., Para [124]-[126] - “The transformation unit is configured to unroll 802 the RNN over a predetermined number of steps. Any of the various approaches known in the art for unrolling (sometimes termed unfolding) an RNN may be used … transformation unit 326 unrolls the RNN over a predetermined number of steps so as to derive 803 a static neural network which represents a portion of the complete unrolled RNN which is mathematically equivalent to the received representation of the RNN. ”), the static network being adapted to perform operations on the fixed size input which are equivalent to the operations performed by the dynamic neural network on its input tensor (see e.g., Para [112][126]-“”); and
implementing a plurality of instances of the static network in hardware for operation on an input tensor split into a sequence, each instance of the static network being arranged to operate on a respective fixed size input of the sequence (see e.g., Para [116][128]-[130]-“ The data processing system further comprises iteration logic 342 which is configured to iteratively apply 805 the derivative neural network to the input sequence and to cause the state outputs from each instance of the derivative neural network (e.g. 404 in FIG. 4) as the state inputs to the next instance of the derivative neural network (e.g. 402 in FIG. 4). The iteration logic may cause each instance of the derivative neural network to be implemented at the accelerator 302 by providing the same derivative neural network to the control logic for implementation at the accelerator each time the current instance of the derivative neural network implemented at the accelerator has completed its processing.”).
Asad does not expressly show input tensor having a variable dimension and split the tensor into a sequence of overlapping fixed size inputs along its variable dimension. However, Hofer teaches the similar feature (see e.g. para [9] – “FIG. 4 shows an example of a moderate to high level of variance relative to the acceptable data series. This moderate to high level of variance may cause at least some training of a stateful RNN to become technologically difficult, time-consuming, and laborious.”). Both Asad and Hofer are directed to model training. Accordingly, it would have been obvious to the skilled artisan before the effective filing date of the claimed invention having Asad and Hofer in front of them to modify the system of Asad to include the above feature. The motivation to combine Asad and Hofer comes from Hofer. Hofer discloses the motivation to process input with variance (see e.g. Abstract para [3]-[9]). This motivation for combination also applies to the remaining claims which depend on this combination.
Asad-Hofer does not expressly show the sequence is of overlapping fixed size inputs along its variable dimension. However, PyTorch teaches the similar feature (see page 1 – “My network is trained with tensors of size BxCx128x128, but I need to verify its image reconstruction performance with images of size 1024x1024. To make the reconstruction smooth, I need to split my input of size BxCx1024x1024 into BxCx128x128 tensors with overlap, which are then fed to the network for reconstruction. “). Both Asad and PyTorch are directed to tensor input and output methods. Accordingly, it would have been obvious to the skilled artisan before the effective filing date of the claimed invention having Asad and PyTorch in front of them to further modify the modified system of Asad to include the above feature. The motivation to combine Asad and PyTorch comes from PyTorch. PyTorch discloses the motivation to split tensors to improve performance (see e.g. page 1). This motivation for combination also applies to the remaining claims which depend on this combination.
Claim 17 is rejected for the similar reasons discussed above with respect to claim 1.
Claim 20 is rejected for the similar reasons discussed above with respect to claim 1.
It is noted that any citation to specific pages, columns, lines, or figures in the prior art references and any interpretation of the references should not be considered to be limiting in any way. “The use of patents as references is not limited to what the patentees describe as their own inventions or to the problems with which they are concerned. They are part of the literature of the art, relevant for all they contain.” In re Heck, 699 F.2d 1331, 1332-33, 216 USPQ 1038, 1039 (Fed. Cir. 1983) (quoting In re Lemelson, 397 F.2d 1006, 1009, 158 USPQ 275, 277 (CCPA 1968)). Further, a reference may be relied upon for all that it would have reasonably suggested to one having ordinary skill the art, including nonpreferred embodiments. Merck & Co. v. Biocraft Laboratories, 874 F.2d 804, 10 USPQ2d 1843 (Fed. Cir.), cert. denied, 493 U.S. 975 (1989). See also Upsher-Smith Labs. v. Pamlab, LLC, 412 F.3d 1319, 1323, 75 USPQ2d 1213, 1215 (Fed. Cir. 2005); Celeritas Technologies Ltd. v. Rockwell International Corp., 150 F.3d 1354, 1361, 47 USPQ2d 1516, 1522-23 (Fed. Cir. 1998).
Conclusion
Any inquiry concerning this communication or earlier communications from the examiner should be directed to PEIYONG WENG whose telephone number is (571)270-1660. The examiner can normally be reached on Mon.-Fri. 8 am to 5 pm.
If attempts to reach the examiner by telephone are unsuccessful, the examiner's supervisor, Matthew Ell, can be reached on (571) 270-3264. The fax phone number for the organization where this application or proceeding is assigned is 571-273-8300.
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://portal.uspto.gov/external/portal. Should you have questions on access to the Private PAIR system, contact the Electronic Business Center (EBC) at 866-217-9197 (toll-free).
/PEI YONG WENG/Primary Examiner, Art Unit 2141
Prosecution Insights