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 .
Claims 1, 5-10, and 13-15 are pending in this application. Claims 1, 5-10, and 13-14 are amended and claims 2-4 and 11-12 are canceled by applicant’s amendment filed 8 April 2026.
Claim Rejections - 35 USC § 112
The following is a quotation of 35 U.S.C. 112(b):
(b) CONCLUSION.—The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the inventor or a joint inventor regards as the invention.
The following is a quotation of 35 U.S.C. 112 (pre-AIA ), second paragraph:
The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the applicant regards as his invention.
Claims 1, 5-10, and 13-15 are rejected under 35 U.S.C. 112(b) or 35 U.S.C. 112 (pre-AIA ), second paragraph, as being indefinite for failing to particularly point out and distinctly claim the subject matter which the inventor or a joint inventor (or for applications subject to pre-AIA 35 U.S.C. 112, the applicant), regards as the invention.
Regarding Claim 1, it recites “wherein the input layer is connected to the excitatory hidden layer and the inhibitory hidden layer via excitatory synapses, respectively” and “wherein the excitatory hidden layer is connected to the excitatory output layer and the inhibitory output layer via excitatory synapses, respectively.” The “respective” connections are unclear because there is no obvious ordering of connections, nor is there a clear distinction among the various possible connections. The input layer contains many excitatory synapses, but the claim does not state which ones are connected to the excitatory hidden layer and which ones are connected to the inhibitory hidden layer. So, the term “respectively” is unclear in this context. Similarly, the excitatory hidden layer contains many excitatory synapses, but the claim does not state which ones are connected to the excitatory output layer and which ones are connected to the inhibitory output layer. Here too, the term “respectively” is confusing and unclear. For the purposes of examination under prior art, the examiner will interpret the present claim to mean that the input layer has at least one connection to the excitatory hidden layer and at least one connection to the inhibitory hidden layer, and that the excitatory hidden layer has at least one connection to the excitatory output layer and at least one connection to the inhibitory output layer.
Regarding Claim 10, it recites limitations substantially similar to those described above for claim 1, so it is indefinite for the same reasons.
Regarding Claims 5-9 and 13-15, they are rejected as being dependent on rejected base claims.
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, 6, 10, and 15 are rejected under 35 U.S.C. 103 as being unpatentable over Chakraborty, Biswadeep, Xueyuan She, and Saibal Mukhopadhyay (“A fully spiking hybrid neural network for energy-efficient object detection,” IEEE Transactions on Image Processing 30 (2021): 9014-9029; hereinafter “Chakraborty”) in view of Dockendorf et al. (U.S. Patent 10,496,922, hereinafter “Dockendorf”), and further in view of Kim et al. (U.S. 2023/0004777, hereinafter “Kim”).
Regarding Claim 1, Chakraborty teaches an image recognition device using a brain-inspired spiking neural network (Abstract and section I, first paragraph), comprising:
an input unit configured to receive an input image made up of at least one or more pixels (fig. 3; section IV. A—an input unit receives an image made up of pixels);
a spiking neural network (SNN) unit configured to recognize the input image, the SNN unit including a plurality of neurons, each provided to correspond to a respective one of pixels of the image and configured to generate a spike signal when a membrane potential state value exceeds a preset threshold value, and synapses connecting the plurality of neurons (fig. 3; section IV. A—the input layer includes one neuron per pixel of the input image); and
an encoding unit configured to encode a neural code to be provided to the plurality of neurons based on luminance of the pixels within the image (fig. 3; section IV. A—intensity {luminance} values of the pixels are encoded as spikes),
wherein the encoding unit performs rate code encoding to determine a firing rate of the spike signal according to the luminance (fig. 3; section IV. A—the encoding assigns firing rates proportional to the intensity {luminance} values of the pixels) {{{, and spike timing code encoding to determine a spike timing according to the luminance}}},
wherein the SNN unit comprises:
an input layer configured to receive the encoded neural code by allocating one neuron for each pixel of the image (fig. 3; section IV. A); and
an output layer (fig. 3; section IV. A. 1);
such that a specific excitatory neuron is induced to selectively react to a specific image (section IV. A. 1—a specific output neuron with the largest number of spikes is selected for a specific input image, which corresponds to a correct label for the image).
Chakraborty does not specifically teach:
wherein the encoding unit performs spike timing code encoding to determine a spike timing according to the luminance;
wherein the SNN unit comprises:
a hidden layer configured to receive signals from a portion of the plurality of neurons constituting the input layer, the hidden layer including an excitatory hidden layer composed of excitatory neurons and an inhibitory hidden layer composed of inhibitory neurons; and
the output layer configured to receive signals from a portion of the neurons constituting the excitatory hidden layer, the output layer including an excitatory output layer composed of excitatory neurons and an inhibitory output layer composed of inhibitory neurons,
wherein the input layer is connected to the excitatory hidden layer and the inhibitory hidden layer via excitatory synapses, respectively,
wherein the excitatory hidden layer and the inhibitory hidden layer are interconnected on a same layer via excitatory synapses and inhibitory synapses,
wherein the excitatory hidden layer is connected to the excitatory output layer and the inhibitory output layer via excitatory synapses, respectively, and
wherein the excitatory output layer and the inhibitory output layer are interconnected on a same layer via excitatory synapses and inhibitory synapses.
However, Dockendorf teaches wherein an SNN unit comprises:
a hidden layer configured to receive signals from a portion of the plurality of neurons constituting the input layer, the hidden layer including an excitatory hidden layer composed of excitatory neurons and an inhibitory hidden layer composed of inhibitory neurons (fig. 5; col. 9, line 55 – col. 10, line 6—a hidden layer includes excitatory neurons 502 and inhibitory neurons 504, and receives signals from neurons in input layer 500); and
the output layer configured to receive signals from a portion of the neurons constituting the excitatory hidden layer, the output layer including an excitatory output layer composed of excitatory neurons and an inhibitory output layer composed of inhibitory neurons (fig. 5; col. 9, line 55 – col. 10, line 6—an output layer includes excitatory neurons 508 and inhibitory neurons 510, and receives signals from neurons in the hidden layer),
wherein the input layer is connected to the excitatory hidden layer and the inhibitory hidden layer via excitatory synapses, respectively (fig. 5; col. 9, line 55 – col. 10, line 6—the input layer is connected directly to excitatory hidden layer 502 through excitatory synapses, and indirectly to inhibitory hidden layer 504 through the excitatory synapses and additional synapses from hidden neurons 502 to inhibitory neurons 504, and also via synapses through the output layer. Additional direct connections from the input neurons to the inhibitory hidden neurons would have been an obvious variation),
wherein the excitatory hidden layer and the inhibitory hidden layer are interconnected on a same layer via excitatory synapses and inhibitory synapses (fig. 5; col. 9, line 55 – col. 10, line 6—additional + and – arrows show synapses connecting excitatory hidden layer 502 to inhibitory hidden layer 504),
wherein the excitatory hidden layer is connected to the excitatory output layer and the inhibitory output layer via excitatory synapses, respectively (fig. 5; col. 9, line 55 – col. 10, line 6—excitatory synapses connect excitatory hidden layer 502 to excitatory output layer 508 and to inhibitory output layer 506), and
wherein the excitatory output layer and the inhibitory output layer are interconnected on a same layer via excitatory synapses and inhibitory synapses (fig. 5; col. 9, line 55 – col. 10, line 6— additional + and – arrows show synapses connecting excitatory output layer 508 to inhibitory output layer 506).
These claimed elements were known in Chakraborty and Dockendorf and could have been combined by known methods with no change in their respective functions. It therefore would have been obvious to a person of ordinary skill in the art at the time of filing of the applicant’s invention to combine the SNN layers of Dockendorf with the SNN and encoding unit of Chakraborty to yield the predictable result of wherein the SNN unit comprises: a hidden layer configured to receive signals from a portion of the plurality of neurons constituting the input layer, the hidden layer including an excitatory hidden layer composed of excitatory neurons and an inhibitory hidden layer composed of inhibitory neurons; and the output layer configured to receive signals from a portion of the neurons constituting the excitatory hidden layer, the output layer including an excitatory output layer composed of excitatory neurons and an inhibitory output layer composed of inhibitory neurons, wherein the input layer is connected to the excitatory hidden layer and the inhibitory hidden layer via excitatory synapses, respectively, wherein the excitatory hidden layer and the inhibitory hidden layer are interconnected on a same layer via excitatory synapses and inhibitory synapses, wherein the excitatory hidden layer is connected to the excitatory output layer and the inhibitory output layer via excitatory synapses, respectively, and wherein the excitatory output layer and the inhibitory output layer are interconnected on a same layer via excitatory synapses and inhibitory synapses, such that a specific excitatory neuron is induced to selectively react to a specific image. One would be motivated to make this combination for the purpose of improving the performance of autonomous systems in real-world environments by increasing their capacity (Dockendorf, col. 1, lines 24-30).
Chakraborty/Dockendorf does not specifically teach wherein the encoding unit performs spike timing code encoding to determine a spike timing according to the luminance. However, Kim teaches an encoding unit that performs rate code encoding to determine a firing rate of the spike signal according to an input value, and spike timing code encoding to determine a spike timing according to the input value (figs. 2 and 3; ¶ [0035] – [0037] and [0048] – [0050]).
All of the claimed elements were known in Chakraborty/Dockendorf and Kim and could have been combined by known methods with no change in their respective functions. It therefore would have been obvious to a person of ordinary skill in the art at the time of filing of the applicant’s invention to combine the simultaneous or sequential firing rate encoding and spike timing encoding of Kim with the encoding unit and luminance values of Chakraborty/Dockendorf to yield the predictable result of wherein the encoding unit performs rate code encoding to determine a firing rate of the spike signal according to the luminance, and spike timing code encoding to determine a spike timing according to the luminance. One would be motivated to make this combination for the purpose of improving artificial intelligence by more accurately mimicking the operation of the human brain, as suggested by Kim, ¶ [0003].
Regarding Claim 6, Chakraborty/Dockendorf/Kim teaches wherein the rate code encoding encodes the neural code by calculating a luminance value of a current pixel relative to an average luminance value of the pixels in the image as the firing rate, such that the firing rate increases as the luminance increases, and wherein the spike timing code encoding encodes the neural code by subtracting a percentage of the luminance value of the current pixel relative to a maximum luminance value from a preset spike reference time, such that the spike timing becomes faster as the luminance increases (Kim, ¶ [0036] – [0038]—the spike rate is encoded to be proportional to the strength of the input signal relative to an average, and the spike timing {temporal coding} encodes the spike time proportional to the strength of the input signal based on a time margin. These operations are equivalent to the calculations of the present claim. Chakraborty teaches that the input intensity values are pixel luminance values in fig. 3 and section IV, A).
Regarding Claim 10, Chakraborty teaches an image recognition method in an image recognition device using a brain-inspired spiking neural network (Abstract and section I, first paragraph). Chakraborty, Dockendorf, and Kim teach the method comprising the operations of the present claim in the same manner as for claim 1, above.
Regarding Claim 15, Chakraborty/Dockendorf/Kim teaches a computer-readable program stored in a computer-readable recording medium configured to perform the image recognition method using a brain- inspired spiking neural network defined in claim 10 (Dockendorf, col. 7, lines 36-54).
Claim 5 is rejected under 35 U.S.C. 103 as being unpatentable over Chakraborty in view of Dockendorf in view of Kim, as applied to claim 1, above, and further in view of Jang, Hyun Jae, et al. (“Distinct roles of parvalbumin and somatostatin interneurons in gating the synchronization of spike times in the neocortex,” Science Advances 6.17 (2020): eaay5333; hereinafter “Jang”).
Regarding Claim 5, Chakraborty/Dockendorf/Kim does not specifically teach wherein the inhibitory neurons include parvalbumin (PV)-expressing inhibitory neurons and somatostatin (SST)-expressing inhibitory neurons. However, Jang teaches wherein inhibitory neurons include parvalbumin (PV)-expressing inhibitory neurons and somatostatin (SST)-expressing inhibitory neurons (p. 1, Introduction. The “Simulation of computational network model” section on pp. 12-13 describes an implementation using a spiking neural network).
All of the claimed elements were known in Chakraborty/Dockendorf/Kim and Jang and could have been combined by known methods with no change in their respective functions. It therefore would have been obvious to a person of ordinary skill in the art at the time of filing of the applicant’s invention to combine the PV and SST inhibitory neurons of Jang with the inhibitory neurons of Chakraborty/Dockendorf/Kim to yield the predictable result of wherein the inhibitory neurons include parvalbumin (PV) expressing inhibitory neurons and somatostatin (SST) expressing inhibitory neurons. One would be motivated to make this combination for the purpose of improving the synchronization of spike timing and firing rates by accounting for the different contributions of PV and SST inhibition (Jang, pp. 8-9, “Discussion” section).
Claims 7-9 and 13-14 are rejected under 35 U.S.C. 103 as being unpatentable over Chakraborty in view of Dockendorf in view of Kim, as applied to claims 1 and 10, and further in view of Wang, Jinling, et al. (“An online supervised learning method for spiking neural networks with adaptive structure,” Neurocomputing 144 (2014): 526-536; hereinafter “Wang”).
Regarding Claims 7 and 13, Chakraborty/Dockendorf/Kim does not specifically teach a learning unit configured to modify synaptic weights by applying a Spike Timing-Dependent Plasticity (STDP) learning rule to synapses between the excitatory neurons such that neurons of the output layer selectively generate spike signals according to the image.
However, Wang teaches a learning unit configured to modify synaptic weights by applying a Spike Timing-Dependent Plasticity (STDP) learning rule to synapses between excitatory neurons such that neurons of an output layer selectively generate spike signals according to an input (section 2.3.3).
All of the claimed elements were known in Chakraborty/Dockendorf/Kim and Wang and could have been combined by known methods with no change in their respective functions. It therefore would have been obvious to a person of ordinary skill in the art at the time of filing of the applicant’s invention to combine the STDP learning of Wang with the SNN and image input of Chakraborty/Dockendorf/Kim to yield the predictable result of a learning unit configured to modify synaptic weights by applying a Spike Timing-Dependent Plasticity (STDP) learning rule to synapses between the excitatory neurons such that neurons of the output layer selectively generate spike signals according to the image. One would be motivated to make this combination for the purpose of improving the training and classification performance of the SNN (Wang, p. 527, last paragraph).
Regarding Claim 8, Chakraborty/Dockendorf/Kim/Wang teaches wherein the learning unit includes a supervised learning unit configured to determine a target neuron of the output layer according to the image, and induce synaptic potentiation or depression by the STDP learning rule through a rise or fall in membrane potential of the target neuron to allow the target neuron to generate the spike signals (Wang, section 2.3.2—supervised learning is applied to the output layer using STDP).
Regarding Claim 9, Chakraborty/Dockendorf/Kim/Wang teaches wherein the learning unit includes an unsupervised learning unit configured to modify the synaptic weights according to the STDP learning rule based on the output spike signals from the output layer according to the image (Wang, section I, last two paragraphs and section 2.3.2—a self-organized unsupervised competitive Hebbian learning method is applied in addition to the supervised learning).
Regarding Claim 14, Chakraborty/Dockendorf/Kim/Wang teaches wherein the learning step comprises:
supervised learning to determine a target neuron of the output layer according to the image, and induce synaptic potentiation or depression by the STDP learning rule through a rise or fall in membrane potential of the target neuron to allow the target neuron to generate the spike signals (Wang, section 2.3.2—supervised learning is applied to the output layer using STDP), or
unsupervised learning to modify the synaptic weights according to the STDP learning rule based on the output spike signals from the output layer according to the image (Wang, section I, last two paragraphs and section 2.3.2—a self-organized unsupervised competitive Hebbian learning method is applied in addition to the supervised learning).
Response to Arguments
Applicant’s arguments with respect to claims 1, 5-10, and 13-15 have been considered but are moot because the new ground of rejection does not rely on any reference applied in the prior rejection of record for any teaching or matter specifically challenged in the argument. The rejections of the amended claims do not rely on Sumbul et al. (U.S. 2019/0042909), Saunders, Daniel J., Hava T. Siegelmann, and Robert Kozma (“Stdp learning of image patches with convolutional spiking neural networks,” 2018 international joint conference on neural networks (IJCNN). IEEE, 2018), or Richert (U.S. 2014/0219497). As detailed above, Chakraborty in view of Dockendorf teaches the recited SNN architecture with an image as input. Kim teaches an SNN that combines rate encoding and timing encoding. Jang and Wang are still relied on to teach some of the limitations of the dependent claims. Note also the new rejections under 35 U.S.C. 112(b) necessitated by the amendment filed 8 April 2026.
Conclusion
Any inquiry concerning this communication or earlier communications from the examiner should be directed to HAL W SCHNEE whose telephone number is (571) 270-1918. The examiner can normally be reached M-F 7:30 a.m. - 6:00 p.m.
Examiner interviews are available via telephone, in-person, and video conferencing using a USPTO supplied web-based collaboration tool. To schedule an interview, applicant is encouraged to use the USPTO Automated Interview Request (AIR) at http://www.uspto.gov/interviewpractice.
If attempts to reach the examiner by telephone are unsuccessful, the examiner’s supervisor, Michael Huntley can be reached at 303-297-4307. The fax phone number for the organization where this application or proceeding is assigned is 571-273-8300.
Information regarding the status of published or unpublished applications may be obtained from Patent Center. Unpublished application information in Patent Center is available to registered users. To file and manage patent submissions in Patent Center, visit: https://patentcenter.uspto.gov. Visit https://www.uspto.gov/patents/apply/patent-center for more information about Patent Center and https://www.uspto.gov/patents/docx for information about filing in DOCX format. For additional questions, contact the Electronic Business Center (EBC) at 866-217-9197 (toll-free). If you would like assistance from a USPTO Customer Service Representative, call 800-786-9199 (IN USA OR CANADA) or 571-272-1000.
/HAL SCHNEE/ Primary Examiner, Art Unit 2129
Prosecution Insights