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 .
Response to Arguments
Applicant’s arguments filed 03/17/2026 (“Remarks”) have been fully considered:
Regarding 101 abstract idea rejection:
Applicant argues: “As an initial matter, categorizing the limitation "applying a transformation to the identified first data item to generate a second data item to form a training dataset containing the first data item, the second data item, and a signal representing the transformation between the first data item and the second data item, wherein applying the transformation comprises at least one of: inverting the identified first data item, or adjusting a contrast of the identified first data item", is inconsistent with U.S.P.T.O. guidance. Specifically, Example 42 of the U.S.P.T.O. examples on subject matter eligibility outlines a claim having the following limitation, "converting, by a content server, the non-standardized updated information into the standardized format", the guidance provided therein states that the converting step is an additional element that should be considered under step 2A prong 2 and step 2B. As such, the limitation of "applying a transformation to the identified first data item to generate a second data item to form a training dataset containing the first data item, the second data item, and a signal representing the transformation between the first data item and the second data item, wherein applying the transformation comprises at least one of: inverting the identified first data item, or adjusting a contrast of the identified first data item", should be considered an additional element… the step of "applying a transformation to the identified first data item to generate a second data item to form a training dataset containing the first data item, the second data item, and a signal representing the transformation between the first data item and the second data item, wherein applying the transformation comprises at least one of: inverting the identified first data item, or adjusting a contrast of the identified first data item" cannot be practically performed in the human mind, as the human mind is not equipped to invert a data item, or adjust contrast, as claimed. At least for these reasons, the above limitation should be considered an additional element, and should be evaluated for a practical application, and/or significantly more, than any alleged abstract idea.”
Examiner disagrees. The limitation of "applying a transformation to the identified first data item to generate a second data item to form a training dataset containing the first data item, the second data item, and a signal representing the transformation between the first data item and the second data item, wherein applying the transformation comprises at least one of: inverting the identified first data item, or adjusting a contrast of the identified first data item", is interpreted as inverting data such as p’=250-p. therefore, depending on the amount of data, it could be done mentally by evaluating the data. The argued example 42, recited a improvement to computer network functionality, whereas the claim seems to improve quality of mathematical model training. In the present claims, there does not seem to recite improvement to computer network or functionality. The claim collects data, identify data satisfying a criterion and apply transformation such as inversion or contrast to generate training data. The generating of transformed training data do not integrate the abstract idea into practical application because it’s only reciting improvement for results such as improving model training or inference without connecting the improvement to a specific technology which computer functions is improved. The recited transformation to generate training data is a generic machine learning technique and uses generic computers as a tool to perform the abstract idea.
As Applicant's disclosure explains, by applying a transformation to gathered data, the neural network can be trained to recognize the rotation that is applied to the image that it gets as an input/trained without a supervision signal. See [0047] of Applicant's published Specification. Training in this manner allows for improved accuracy of inference over prior models. See [0004] of Applicant's published Specification. The limitation "applying a transformation to the identified first data item to generate a second data item to form a training dataset containing the first data item, the second data item, and a signal representing the transformation between the first data item and the second data item, wherein applying the transformation comprises at least one of: inverting the identified first data item, or adjusting a contrast of the identified first data item" evidences a specific improvement to the functioning of a computer, or to a technology or technical field. Specifically, the limitation, above, represents an improvement to the training of machine learning models resulting in higher accuracy inferencing. As such the courts have found that an improvement in the functioning of a computer, or an improvement to other technology or technical field, are indicative of either, or both of, a practical application, or significantly more than, an abstract idea.
Examiner disagrees. As explained above, the claims do not recite a specific technological improvement to computer functionality. The claim recites applying known image transformation to training data and training the model using resulting dataset. Improving the accuracy of a model without reciting improvement to a computer capabilities do not render the claim eligible. For at least the above reasons, 101 rejection is maintained.
Regarding 103 abstract idea rejection:
Regarding applicant's arguments filed with respect to the prior art rejections have been fully considered but they are moot. Applicant has amended the claims to recite new combinations of limitations. Please see below for new grounds of rejection, necessitated by Amendment.
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 1-20 are rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more. The analysis of the claims’ subject matter eligibility will follow the 2019 Revised Patent Subject Matter Eligibility Guidance, 84 Fed. Reg. 50-57 (January 7, 2019) (“2019 PEG”).
With respect to claim 1.
Claim 1 is rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more.
Step 1: Is the claim to a process, machine, manufacture, or composition of matter? Yes—claim 1 recites a method, which is a process.
Step 2A, prong one: Does the claim recite an abstract idea, law of nature or natural phenomenon? Yes—the limitations identified below each, under its broadest reasonable interpretation, covers mental processes abstract idea grouping (concepts performed in the human mind (including an observation, evaluation, judgment, opinion)), see MPEP 2106.04(a)(2), subsection III and the 2019 PEG, but for the recitation of generic computer components:
“identifying a first data item from among the collected sensor data when the first data item is determined to satisfy a criterion; applying a transformation to the identified first data item to generate a second data item to form a training dataset containing the first data item, the second data item, and a signal representing the transformation between the first data item and the second data item, wherein applying the transformation comprises at least one of: inverting the identified first data item, or adjusting a contrast of the identified first data item;”: (Mental processes- concept of observation and evaluation of selecting data based on criteria and applying transformation to the data or mathematical concept of image transformation).
Step 2A, prong two: Does the claim recite additional elements that integrate the judicial exception into a practical application? No—the judicial exception is not integrated into a practical application.
“receiving, from one or more server computers through a communication network, an edge model;” involves the mere gathering of data, which is insignificant extra-solution activity. See MPEP § 2106.05(g).
“collecting sensor data acquired by a sensor on a vehicle;” involves the mere gathering of data, which is insignificant extra-solution activity. See MPEP § 2106.05(g).
“training with respect to the edge model on the training dataset” and also the use of “one or more server computers through a communication network”, “a sensor on a vehicle”: Merely reciting the words “apply it” (or an equivalent) with the judicial exception, or merely including instructions to implement an abstract idea on a computer, or merely using a computer as a tool to perform an abstract idea, as discussed in MPEP § 2106.05(f).
“transmitting first data representing the trained edge model to the one or more server computers though the communication network.” involves the mere gathering of data, which is insignificant extra-solution activity. See MPEP § 2106.05(g).
The generic computer components in these steps are recited at a high-level of generality (i.e., as a generic computer component performing a generic computer function) such that it amounts no more than mere instructions to apply the exception using a generic computer component. Accordingly, this additional element does not integrate the abstract idea into a practical application because it does not impose any meaningful limits on practicing the abstract idea. The claim is directed to an abstract idea.
Step 2B: Does the claim recite additional elements that amount to significantly more than the judicial exception? No—there are no additional limitations beyond the mental processes identified above. The limitation treated above, are directed to the well-understood, routine, and conventional activity of storing and retrieving information in memory. See MPEP § 2106.05(d)(II); Versata Dev. Group, Inc. v. SAP Am., Inc., 793 F.3d 1306, 1334, 115 USPQ2d 1681, 1701 (Fed. Cir. 2015). It also includes limitations that Merely reciting the words “apply it” (or an equivalent) with the judicial exception, or merely including instructions to implement an abstract idea on a computer, or merely using a computer as a tool to perform an abstract idea, as discussed in MPEP § 2106.05(f). The additional element is insignificant application, which is similar to examples of activities that the courts have found to be insignificant extra-solution activity, in accordance with MPEP 2106.05(g), Insignificant Extra-Solution Activity. Mere instructions to apply an exception using a generic computer component cannot provide an inventive concept.
Thus, considering the additional elements individually and in combination and the claims as a whole, the additional elements do not provide significantly more than the abstract idea. This claim is not patent eligible.
Claim 2.
Step 1: A method, as above.
Step 2A Prong 1: The mental process of claim 1 applies to claim 2.
Step 2A Prong 2, Step 2B: claim recites that “receiving, from the one or more server computers through the communication network, second data that represents a model that is trained with aggregated model information from other edge models; and updating the edge model based on the second data”, involves the mere gathering of data, which is insignificant extra-solution activity. See MPEP § 2106.05(g). This judicial exception is not integrated into a practical application. Mere recitation of generic computer components neither integrates the judicial exception into a practical application nor provides an inventive concept.
Claim 3.
Step 1: A method, as above.
Step 2A Prong 1: The mental process of claim 1 applies to claim 3.
Step 2A Prong 2, Step 2B: claim recites that “wherein the training with respect to the edge model comprises training a copy of the received edge model.” Adding the words “apply it” (or an equivalent) with the judicial exception, or mere instructions to implement an abstract idea on a computer, e.g., a limitation indicating that a particular function such as creating and maintaining electronic records is performed by a computer, as discussed in Alice Corp., 573 U.S. at 225-26, 110 USPQ2d at 1984 (see MPEP § 2106.05(f)); This judicial exception is not integrated into a practical application. Mere recitation of generic computer components neither integrates the judicial exception into a practical application nor provides an inventive concept.
Claim 4.
Step 1: A method, as above.
Step 2A Prong 1: The mental process of claim 1 applies to claim 4.
Step 2A Prong 2, Step 2B: The claim recites that “obtaining, as the first data, a gradient between the edge model prior to the training and the edge model subsequent to the training;” involves the mere gathering of data, which is insignificant extra-solution activity. See MPEP § 2106.05(g). This judicial exception is not integrated into a practical application. Mere recitation of generic computer components neither integrates the judicial exception into a practical application nor provides an inventive concept.
Claim 5.
Step 1: A method, as above.
Step 2A Prong 1: The mental process of claim 1 applies to claim 5.
Step 2A Prong 2, Step 2B: The claim recites that “obtaining, as the first data, a gradient between the received edge model and the copy of the edge model that is updated by the training;” involves the mere gathering of data, which is insignificant extra-solution activity. See MPEP § 2106.05(g). This judicial exception is not integrated into a practical application. Mere recitation of generic computer components neither integrates the judicial exception into a practical application nor provides an inventive concept.
Claim 6 .
Step 1: A method, as above.
Step 2A Prong 1: The claim recites that “wherein the applying the transformation comprises rotating the first data item”: This limitation merely specifies mental processes- concept of observation and evaluation.
Step 2A Prong 2, Step 2B: This judicial exception is not integrated into a practical application. Mere recitation of generic computer components neither integrates the judicial exception into a practical application nor provides an inventive concept.
Claim 7.
Step 1: A method, as above.
Step 2A Prong 1: The mental process of claim 1 applies to claim 7.
Step 2A Prong 2, Step 2B: claim recites that “wherein the training on the training dataset comprises training without human annotation.” Adding the words “apply it” (or an equivalent) with the judicial exception, or mere instructions to implement an abstract idea on a computer, e.g., a limitation indicating that a particular function such as creating and maintaining electronic records is performed by a computer, as discussed in Alice Corp., 573 U.S. at 225-26, 110 USPQ2d at 1984 (see MPEP § 2106.05(f)); This judicial exception is not integrated into a practical application. Mere recitation of generic computer components neither integrates the judicial exception into a practical application nor provides an inventive concept.
Claims 8-14
Step 1: The claims recite a computing device; therefore, they fall into the statutory category of machines.
Step 2A Prong 1: The claims recite the same mental processes as claims 1-7, respectively.
Step 2A Prong 2: This judicial exception is not integrated into a practical application. Claims 8-14 recite generic computer components, namely “a memory storing instructions; and
a processor configured to execute the instructions”. As before, the mere recitation that the method is to be performed on a generic computer amounts to a mere instruction to apply the exception on the computer. See MPEP § 2106.05(f). With that exception, the analysis mirrors that of claims 1-7, respectively.
Step 2B: The claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception. The analysis, with the one exception noted above, mirrors that of claims 1-7, respectively.
Claims 15-20
Step 1: The claims recite a non-transitory computer readable medium; therefore, they fall into the statutory category of machines.
Step 2A Prong 1: The claims 15-20 recite the same mental processes as claims 1-6, respectively.
Step 2A Prong 2: This judicial exception is not integrated into a practical application. Claims 15-20 recite generic computer components, namely “non-transitory computer-readable medium storing instructions, the instructions comprising: one or more instructions that, when executed by one or more processors of a device”. As before, the mere recitation that the method is to be performed on a generic computer amounts to a mere instruction to apply the exception on the computer. See MPEP § 2106.05(f). With that exception, the analysis mirrors that of claims 1-6, respectively.
Step 2B: The claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception. The analysis, with the one exception noted above, mirrors that of claims 1-6, respectively.
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.
Claim(s) 1-5, 8-12 and 15-19 is/are rejected under 35 U.S.C. 103 as being unpatentable over Sprague et al. (US 20200334524 A1) in view of Karras et al. (US 20210383241 A1).
Regarding claim 1.
Sprague teaches a method, implemented by programmed one or more processors, comprising: receiving, from one or more server computers through a communication network, an edge model (see ¶ 46, “Each of the device 122 may store a model (e.g. machine-learned network) that is trained by a large number (hundreds, thousands, millions, etc.) of devices 122 with each device 122 holding a set of training data without sharing data sets. Each device 122 may be configured to training a pre-agreed model with gradient descent learning for a respective piece of training data, only sharing learnt parameters of the model with the rest of the network. The device 122 is configured to acquire different training data than other devices that are training the model. In addition, at least one transmission between the device and a parameter server may occur asynchronously with respect to the other devices that are training the model. The devices 122 are configured to over or under sample acquired data. When over sampling the data, the devices 122 may reuse the data”, also see ¶ 53, “Similar to receiving locally trained model parameters, each parameter server transmits trained model parameters to the master parameter server and received back master trained model parameters.”, also see ¶ 42);
collecting sensor data acquired by a sensor on a vehicle (see p 76, “The worker device 122 may use the same local training data or may update the training data with newly collected sensor data. The training data may be weighted by age or may be cycled out by the device. For example, data older than a day, month, or year, may be retired and no longer used for training purposes. Data may also be removed or deleted by a user or automatically by the device. Additional data may be added to the training data set as the data is collected. In an embodiment, the worker device 122”, also see ¶ 80, “the device (both worker and parameter server) may be smartphones, navigation devices, vehicle systems, etc. Each of the worker devices 122 may include a sensor or input interface that collects data. Example of sensors may include a camera, LIDAR, radar, microphone, etc.”, also ¶ 85);
identifying a first data item from among the collected sensor data when the first data item is determined to satisfy a criterion (see ¶ 61, “At act A120, the worker device 122 selects a first set of data instances from the acquired data instances as a function of a threshold value received from a parameter server. There are two different scenarios for selection of the data instances.”, also see ¶ 75, “At act A160, the worker device 122 selects another set of data instances to be used as training data. The quantity of the data instances in the local training data is regulated by either the original threshold value or if applicable, an updated threshold value received from the parameter server 125.”);
training with respect to the edge model on the training dataset (see 77, “At act A170, the worker device 122 retrains the model using the local training data and the third parameter. The model is trained similarly to the act A130. The difference for each iteration is a different starting point for one or more of the parameters in the model. The central parameter vector that is received may be different than the local parameter vector generated by the device in A130.”);
and transmitting first data representing the trained edge model to the one or more server computers though the communication network (see ¶ 78, “the worker device 122 transmits the fourth parameter of the updated trained model to the parameter server 125. The worker device 122 receives a fifth parameter from the parameter server 125. The process repeats for a number of iteration until the parameters converge or a predetermined number of iteration is reached.”).
Sprague do not specifically teach applying a transformation to the identified first data item to generate a second data item to form a training dataset containing the first data item, the second data item, and a signal representing the transformation between the first data item and the second data item, wherein applying the transformation comprises at least one of: inverting the identified first data item, or adjusting a contrast of the identified first data item.
Karras teaches applying a transformation to the identified first data item to generate a second data item to form a training dataset containing the first data item, the second data item, and a signal representing the transformation between the first data item and the second data item, wherein applying the transformation comprises at least one of: inverting the identified first data item, or adjusting a contrast of the identified first data item (see figure 2, and ¶ 46, “FIG. 2D illustrates transformations using an invertible augmentation operator, in accordance with an embodiment. In contrast with the non-invertible rotation transformation shown in FIG. 2C, the rotation transformation can be made invertible by changing the probability from uniform to a non-uniform. The augmentation unit 200 processes generated data 211 and example output data 213 using an invertible rotation transformation to produce augmented generated data 220 and augmented example output data 222, respectively. The probability of performing an augmentation is set to 80%, so that each of the four rotations is 20%, 20%, 20%, 20% for 0°, 90°, 180°, 270°, respectively. Because no augmentation occurs at 20% probability, the resulting probability for rotating by 0° is 40%.”, also see ¶ 53, “each transformation unit 300 may be configured to perform any one of N transformations. In another embodiment, each transformation unit 300 is configured to perform a specific one of the N transformations. In an embodiment, the N transformations may be grouped into 6 categories: pixel blitting (x-flips, 90° rotations, integer translation), more general geometric transformations, color transforms, image-space filtering, additive noise, and cutout. In an embodiment, during training, each image is shown to the discriminator 110 after being processed by the augmentation unit 200 using a pre-defined set of transformations in a fixed order. Because the augmentations are also used when training the generator 100, the augmentations should be differentiable for backpropagation when computing the generator parameter updates.”).
Both Sprague and Karras pertain to the problem of model training using sensor data, thus being analogous. It would have been obvious to one skilled in the art before the effective filing date of the claimed invention to combine Sprague and Karras to teach the above limitations. The motivation for doing so would be “Training GANs using too little example data typically leads to discriminator overfitting, causing training to diverge and produce poor results. An adaptive discriminator augmentation mechanism is used that significantly stabilizes training with limited data providing the ability to train high-quality GANs. An augmentation operator is applied to the distribution of inputs to a discriminator used to train a generator, representing a transformation that is invertible to ensure there is no leakage of the augmentations into the images generated by the generator. Reducing the amount of training data that is needed to achieve convergence has the potential to considerably help many applications and may the increase use of generative models in fields such as medicine.” (see Karras abstract).
Regarding claim 2.
Sprague and Karras teaches the method according to claim 1,
Sprague further teaches further comprising: receiving, from the one or more server computers through the communication network, second data that represents a model that is trained with aggregated model information from other edge models; and updating the edge model based on the second data (see ¶ 75, “At act A160, the worker device 122 selects another set of data instances to be used as training data. The quantity of the data instances in the local training data is regulated by either the original threshold value or if applicable, an updated threshold value received from the parameter server 125.”, also see ¶ 82-84, aggregation and training local model with updated central parameters).
Regarding claim 3.
Sprague and Karras teaches the method according to claim 1,
Sprague further teaches wherein the training with respect to the edge model comprises training a copy of the received edge model (see ¶ 29, “Using the labeled objects, each worker device may train a locally stored model using a classification technique. Parameters for the locally trained model are transmitted by each of the worker devices to a parameter server. The transmission are quasi-synchronous. For example, each worker device may transmit local parameters after training a copy of a local model on a number of data instances.”, also see ¶ 45).
Regarding claim 4.
Sprague and Karras teaches the method according to claim 1,
Sprague further teaches further comprising obtaining, as the first data, a gradient between the edge model prior to the training and the edge model subsequent to the training (see ¶ 34, “Gradient descent is used to minimize the error functions. Given a function defined by a set of parameters, gradient descent starts with an initial set of parameter values and iteratively moves toward a set of parameter values that minimize the function. The iterative minimization is based on a function that takes steps in the negative direction of the function gradient. A search for minimizing parameters starts at any point and allows the gradient descent algorithm to proceed downhill on the error function towards a best outcome. Each iteration updates the parameters that yield a slightly different error than the previous iteration. A learning rate variable is defined that controls how large of a step that is taken downhill during each iteration.”, also see ¶ 46, ¶ 68, and ¶ 81).
Regarding claim 5.
Sprague and Karras teaches the method according to claim 3,
Sprague further teaches further comprising obtaining, as the first data, a gradient between the received edge model and the copy of the edge model that is updated by the training (see ¶ 81, “A worker device first trains the locally stored model through gradient descent in a pre-arranged fashion (fixed or flexible number of epochs) and then sends the trained parameters to the device housing the process representing the parameter server 125. The parameter server 125 calculates an updated parameter and immediately sends the updated parameter back to the respective device. The parameter server 125 does not wait for additional devices to respond. Upon receipt of the updated parameters, the process representing that worker unit continues its training of the model locally using local data.”).
Claims 8-12 recites a computing device to perform the method recited in claims 1-5. Therefore the rejection of claims 1-5 above applies equally here. Sprague also teaches the addition elements of claim 8 not recited in claim 1 comprising: a memory storing instructions; and a processor configured to execute the instructions (see figure 1, memory and processor).
Claims 15-19 recites a non-transitory computer-readable medium storing instructions to perform the method recited in claims 1-5. Therefore the rejection of claims 1-5 above applies equally here. Sprague also teaches the addition elements of claim 15 not recited in claim 1 comprising: one or more instructions that, when executed by one or more processors of a device, cause the one or more processors (see figure 1, memory and processor).
Claim(s) 6-7, 13-14 and 20 is/are rejected under 35 U.S.C. 103 as being unpatentable over Sprague et al. (US 20200334524 A1) in view of Karras et al. (US 20210383241 A1) in further in view of Goto et al. (US 20200110994 A1).
Regarding claim 6.
Sprague and Karras teaches the method according to claim 1,
Sprague and Karras do not teach the limitation of claim 6.
Goto teaches wherein the applying the transformation further comprises rotating the first data item (see ¶ 26, “At operations 240-250, the system enters a loop in which affine or other transformations are performed on the dataset. At operation 240, the system may randomly transform the m.sup.th dataset in an affine manner such that features of the dataset are maintained. For example, affine transformation may preserve certain features such as points, straight lines, and planes, as well as sets of parallel lines and ratios of distances between points on a straight line. However, other features such as angles between lines or distances between points may not be preserved. Examples of affine transformations may include but are not limited to translation, scaling, reflection, rotation, shear mapping, enlargement, reduction, similarity transformation, or any combination thereof. At operation 245, the augmented sample index (1) is incremented. At operation 250, if the augmented sample index (1) is less than the magnitude of augmentation for class c (L.sub.c), operations 240 and 245 are repeated.”, also see ¶ 35 and figures 2-4).
Sprague, Karras and Goto pertain to the problem of model training, thus being analogous. It would have been obvious to one skilled in the art before the effective filing date of the claimed invention to combine Sprague, Karras and Goto to teach the above limitations. The motivation for doing so would be “dataset comprising a plurality of classes is obtained for training a neural network. Prior to initiation of training, the dataset may be augmented by performing affine transformations of the data in the dataset, wherein the amount of augmentation is determined by a data augmentation variable. The neural network is trained with the augmented dataset. A training loss and a difference of class accuracy for each class is determined. The data augmentation variable is updated based on the total loss and class accuracy for each class. The dataset is augmented by performing affine transformations of the data in the dataset according to the updated data augmentation variable, and the neural network is trained with the augmented dataset.” (see Goto abstract).
Regarding claim 7.
Sprague and Karras teaches the method according to claim 1,
Sprague and Karras do not teach the limitation of claim 7.
Goto teaches wherein the training on the training dataset comprises training without human annotation (see ¶ 26, “At operations 240-250, the system enters a loop in which affine or other transformations are performed on the dataset. At operation 240, the system may randomly transform the m.sup.th dataset in an affine manner such that features of the dataset are maintained. For example, affine transformation may preserve certain features such as points, straight lines, and planes, as well as sets of parallel lines and ratios of distances between points on a straight line. However, other features such as angles between lines or distances between points may not be preserved. Examples of affine transformations may include but are not limited to translation, scaling, reflection, rotation, shear mapping, enlargement, reduction, similarity transformation, or any combination thereof. At operation 245, the augmented sample index (1) is incremented. At operation 250, if the augmented sample index (1) is less than the magnitude of augmentation for class c (L.sub.c), operations 240 and 245 are repeated.”, also see ¶ 35 and figures 2-4).
Sprague, Karras and Goto pertain to the problem of model training, thus being analogous. It would have been obvious to one skilled in the art before the effective filing date of the claimed invention to combine Sprague, Karras and Goto to teach the above limitations. The motivation for doing so would be “dataset comprising a plurality of classes is obtained for training a neural network. Prior to initiation of training, the dataset may be augmented by performing affine transformations of the data in the dataset, wherein the amount of augmentation is determined by a data augmentation variable. The neural network is trained with the augmented dataset. A training loss and a difference of class accuracy for each class is determined. The data augmentation variable is updated based on the total loss and class accuracy for each class. The dataset is augmented by performing affine transformations of the data in the dataset according to the updated data augmentation variable, and the neural network is trained with the augmented dataset.” (see Goto abstract).
Claims 13-14 recites a computing device to perform the method recited in claims 6-7. Therefore the rejection of claims 6-7 above applies equally here.
Claim 20 recites a non-transitory computer-readable medium storing instructions to perform the method recited in claim 6. Therefore the rejection of claim 6 above applies equally here.
Related prior arts:
SANTAMARIA et al. (US 20240346380 A1) teaches collected sensor data and simulated sensor data created by transforming collected sensor data are used to train a machine learning model (MLM), the MLM is then deployed on an integrated circuit chip of an embedded device, live sensor data received by the embedded device is then either transformed and input to the MLM or input to the MLM without transformation, and the MLM then performs a prediction by, for example, recognizing a gesture made by the user of the embedded device.
Bryant et al. (US 20210264526 A1) teaches distributed ledger for use with vehicle sensor data and usage based systems. These usage based systems may be opt-in programs wherein a driver volunteers to install one or more vehicle sensors configured to monitor various characteristics associated with a vehicle (e.g., speed, acceleration, cornering, braking, path, etc.) that are utilized to develop a profile including information quantifying, for example, a driver's habits.
BOND et al. (US 20220284662 A1) teaches receive first data from a first set of sensors arranged in a first configuration. The instructions configure the processor to transform the first data to a second data to train a model to recognize third data captured by a second set of sensors arranged in a second configuration. The second configuration is different than the first configuration. The instructions configure the processor to train the model based on the second set of sensors sensing the second data to recognize the third data captured by the second set of sensors arranged in the second configuration.
Hu et al. (US 20210056306 A1) teaches resulted in improved capabilities in object identification and analysis. Machine learning has been used as a tool for detecting objects in image data for purposes of such analysis. In order to train machine learning models to perform object identification, however, conventional approaches require a significant amount of labeled training data, where supervised training data includes ground truth data. Creating this training data can be a long and complicated process, which may be too expensive for various uses and may result in an insufficient amount of training data.
Conclusion
Applicant's amendment necessitated the new ground(s) of rejection presented in this Office action. Accordingly, THIS ACTION IS MADE FINAL. See MPEP § 706.07(a). 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 IMAD M KASSIM whose telephone number is (571)272-2958. The examiner can normally be reached 10:30AM-5:30PM, M-F (E.S.T.).
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 J. 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.
/IMAD KASSIM/Primary Examiner, Art Unit 2129
Prosecution Insights