CTFR 18/136,687 CTFR 100919 Notice of Pre-AIA or AIA Status 07-03-aia AIA 15-10-aia The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA. Priority 02-27 AIA Acknowledgment is made of applicant’s claim for foreign priority under 35 U.S.C. 119 (a)-(d). The certified copy has been filed in parent Application No. KR10-2022-0145335 , filed on 11/03/2022 . 12-151 AIA 26-51 12-51 Status of Claims Claims 1, 5, 11, and 15 were amended. Claims 2 -4 and 12 – 14 have been cancelled. Claims 1, 5 – 11, 15 – 20 are pending and are examined herein. Claims 1, 5 – 11, 15 – 20 are rejected under 35 U.S.C. 112(b). Claims 1, 5 – 11, 15 – 20 are rejected under 35 U.S.C. 101. Claims 1, 5 – 11, 15 – 20 are rejected under 35 U.S.C. 103. Response to Amendment The amendment filed March 17 th , 2026 has been entered. Claims 1, 5, 11, and 15 were amended. Claims 2 -4 and 12 – 14 have been cancelled. Claims 1, 5 – 11, 15 – 20 are pending and are examined herein. Applicant’s amendments to the claims have overcome each and every 112(b) rejection previously set forth in the Non-Final Rejection Office Action mailed December 17 th , 2025. However, new 112(b) rejection has been introduced to the current amendment. Response to Arguments 07-37 AIA Applicant's arguments filed March 17 th , 2026 regarding the 35 U.S.C. 101 rejection have been fully considered but they are not persuasive. Applicant argues, on page 7 – 8, that the amended limitations do not recite a judicial exception because they recite specific data preparation and model training operations. The amended limitation are directed to collecting, comparing, organizing, and labeling information regarding objects observed at different time points. The claimed first table, second table, field, and token are data representations used to record and classify object information. The claimed token merely represents the result of determining that an object present at a current time point is absent at a future time point. Therefore, the amended limitations recite a mental process because they involve observation, evaluation, judgment, and classification of information. A person reviewing sequential images could identify objects present in a first image, identify objects present in a later image, compare the two set of objects, and label an object as missing or disappeared. MPEP § 2106 explains that mental processes include observations, evaluations, judgments, and opinions, and that the courts do not distinguish between mental processes performed entirely in the human mind and mental processes performed with pen and paper or on a computer. Applicant further argue that the claim is performed by a controller on training images but that does not remove the claim from the mental process grouping. The use of a generic controller to perform the same information processing steps does not change the character of the claim. The claim does not recite a specific image processing algorithm, a specific training algorithm, a specific model parameter update method, or a specific technical improvement to the model architecture . Applicant further argues, on page 8 – 9, that the amended claims integrate the alleged abstract idea into a practical application as they improve path prediction technology for vehicles. The additional elements in present disclosure do not integrate the abstract idea into a practical application. The sensor merely obtains training images, which is data gathering, and the controller is recited generically. The real environment and path prediction context merely limits the abstract idea to a particular filed of use. The path prediction model is also recited generically and is not claimed with any particular model architecture, training loss, parameter update method, computational efficiency improvement, or specific vehicle control operation. The alleged improvement by applicant is that the model may better account for objects that disappear from a future image. However, the claim does not recite a specific technical mechanism for improving the functioning of the computer, sensor, controller, vehicle, or machine learning model itself. Rather, the claim recites preparing labeled information and using that labeled information in training. An asserted improvement in prediction accuracy, without a claimed technical mechanism for achieving that improvement, is not sufficient to integrate the abstract idea into a practical application. MPEP § 2106 explains that merely adding generic computer components or implementing an abstract idea on a generic computer does not automatically overcome an eligibility rejection. Applicant further argues, on page 8 – 10, that the amended limitations provide an inventive concept. The additional elements, considered individually and as an ordered combination, do not amount to significantly more than the abstract idea. The claims recite generic additional elements, such as sensor, controller, data structures, token, and they merely implement the abstract information organization and labeling concept using ordinary computer components. The claims do not recite an unconventional arrangement of hardware, a specific improvement to computer functionality, or a specific technological operation beyond using the tokenized information in model training. Thus, claims 1, 5-11, and 15-20 remain rejected under 35 U.S.C. § 101. Applicant’s arguments, see pages 9-12, filed March 17 th , 2026 with respect to the rejection(s) of claim(s) 1-20 under 35 U.S.C. 103 have been fully considered and but they are not persuasive. Further details on the 103 rejection with respect to amended limitations are provided in the 103 rejection section of the current office action . Claim Rejections - 35 USC § 112 07-30-02 AIA 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. 07-34-01 Claims 1, 5 – 11, 15 – 20 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. Claim 1 recites “generate a second table in which dynamic objects in a training image at a future time point are recorded.” and further recites “assign, to a field of a first dynamic object in the second table, a token indicating that the first dynamic object is a dynamic object that disappears at the future time point when the first dynamic object in the training image at the current time point disappears from the training image at the future time point.” The scope of the limitation is unclear because the second table is defined as recording dynamic objects in the training image at the future time point, but the claim also requires assigning a token to a field of the first dynamic object in the second table when the first dynamic object disappears from the training image at the future time point. If the first dynamic object disappears from the future training image, it is unclear whether the first dynamic object is included in the second table, whether the second table includes a placeholder entry for an object absent from the future image, or whether the second table records both objects present in the future image and objects from the current image that are absent from the future image. Therefore, the metes and bounds of the claim are unclear. Claims 5 – 10 depends on claim 1. They do not resolve the issue of indefiniteness and are rejected with the same rationale. Claim 5 recites the limitation “the controller is configured to train a second dynamic object that is not present in a training image…” The scope of this limitation is unclear because the claim recites training a dynamic object itself. A dynamic object appearing in an image is not itself trained. It is unclear whether the claim requires training the path prediction model to classify the second dynamic object as newly appearing, assigning or using data associated with the second dynamic object as training data. Therefore, the metes and bounds of the claim are unclear. Claim 6 is dependent on claim 5. It does not resolve the issue of indefiniteness and is rejected with the same rationale. Claims 11, 15 – 20 recite substantially similar subject matter to claims 1, 5-10 respectively and are rejected with the same rationale, mutatis mutandis . Claim Rejections - 35 USC § 101 07-04-01 AIA 07-04 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, 5 - 11, 15 - 20 are rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more. MPEP § 2109(III) sets out steps for evaluating whether a claim is drawn to patent-eligible subject matter. The analysis of claims 1, 5 - 11, 15 - 20, in accordance with these steps, follows. Step 1 Analysis: Step 1 is to determine whether the claim is directed to a statutory category (process, machine, manufacture, or composition of matter. Claims 1, 5 – 10 are an apparatus for path prediction model, meaning that it is directed to the statutory category of machine. Claims 11, 15 – 20 are directed to a method for training a path prediction model, which can be an article of process. Step 2A Prong One, Step 2A Prong Two, and Step 2B Analysis: Step 2A Prong One asks if the claim recites a judicial exception (abstract idea, law of nature, or natural phenomenon). If the claim recites a judicial exception, analysis proceeds to Step 2A Prong Two, which asks if the claim recites additional elements that integrate the abstract idea into a practical application. If the claim does not integrate the judicial exception, analysis proceeds to Step 2B, which asks if the claim amounts to significantly more than the judicial exception. If the claim does not amount to significantly more than the judicial exception, the claim is not eligible subject matter under 35 U.S.C. 101. Regarding claim 1 , the following claim elements are abstract ideas: generate a first table in which dynamic objects in a training image at a current time point are recorded, (Recording observed object information at a time point is practical to perform in the human mind under its broadest reasonable interpretation aside from the recitation of generic computer components or by a human using a pen and paper.) generate a second table in which dynamic objects in a training image at a future time point are recorded, (Recording observed object information at a time point is practical to perform in the human mind under its broadest reasonable interpretation aside from the recitation of generic computer components or by a human using a pen and paper.) assign, to a field of a first dynamic object in the second table, a token indicating that the first dynamic object is a dynamic object that disappears at the future time point when the first dynamic object in the training image at the current time point disappears from the training image at the future time point, (Comparing current and future object information and labeling an object as disappeared is practical to perform in the human mind under its broadest reasonable interpretation aside from the recitation of generic computer components or by a human using a pen and paper.) The following claim elements are additional elements which, taken alone or in combination with the other additional elements, do not integrate the judicial exception into a practical application nor amount to significantly more than the judicial exception: An apparatus for path prediction model training, the apparatus comprising: (This falls under mere instructions to apply an exception. See MPEP § 2106.05(f). Therefore, this does not amount to significantly more than the judicial exception.) a sensor configured to obtain a time series of training images in a real environment; (This is mere data gathering, an insignificant extra solution activity, which does not integrate the judicial exception into a practical application. The broadest reasonable interpretation of this claim is storing information in memory, which is a well-understood, routine conventional activity. See MPEP § 2106.05(d)(II)(iv). Therefore, this does not amount to significantly more than the judicial exception.) and a controller configured to train a path prediction model based on dynamic objects in the time series of training images, wherein the controller is configured to train the path prediction model (This falls under mere instructions to apply an exception. See MPEP § 2106.05(f). Therefore, this does not amount to significantly more than the judicial exception.) and train the path prediction model using the token. (This falls under mere instructions to apply an exception. See MPEP § 2106.05(f). Therefore, this does not amount to significantly more than the judicial exception.) Regarding claim 5 , the rejection of claim 1 is incorporated herein. Further, claim 5 recites the following additional element: wherein the controller is configured to train a second dynamic object that is not present in a training image at a past time point as a dynamic object newly appearing in the training image at the current time point when the second dynamic object newly appears in the training image at the current time point. (This falls under mere instructions to apply an exception. See MPEP § 2106.05(f). Therefore, this does not amount to significantly more than the judicial exception.) Regarding claim 6 , the rejection of claim 5 is incorporated herein. Further, claim 6 recites the following abstract idea: determine a value indicating that the second dynamic object is the dynamic object newly appearing at the current time point (This is practical to perform in the human mind under its broadest reasonable interpretation aside from the recitation of generic computer components or by a human using a pen and paper.) Claim 6 further recites following additional elements: wherein the controller is configured to … and train the path prediction model using the value. (This falls under mere instructions to apply an exception. See MPEP § 2106.05(f). Therefore, this does not amount to significantly more than the judicial exception.) Regarding claim 7 , the rejection of claim 1 is incorporated herein. Further, claim 7 recites the following additional element: wherein the path prediction model is a transformer network. (This falls under mere instructions to apply an exception. See MPEP § 2106.05(f). Therefore, this does not amount to significantly more than the judicial exception.) Regarding claim 8 , the rejection of claim 7 is incorporated herein. Further, claim 8 recites the following additional element: wherein the transformer network is configured to train a location of each of the dynamic objects at a future time point based on an input vector of each of the dynamic objects at a past time point and an input vector of each of the dynamic objects at a current time point. (This falls under mere instructions to apply an exception. See MPEP § 2106.05(f). Therefore, this does not amount to significantly more than the judicial exception.) Regarding claim 9 , the rejection of claim 1 is incorporated herein. Further, claim 9 recites the following additional element: the dynamic objects in the time series of training images are vehicles, (This falls under mere instructions to apply an exception. See MPEP § 2106.05(f). Therefore, this does not amount to significantly more than the judicial exception.) and the controller is configured to extract feature information about each of the vehicles in the time series of training images as an input vector for the path prediction model. (This is mere data gathering, an insignificant extra solution activity, which does not integrate the judicial exception into a practical application. The broadest reasonable interpretation of this claim is storing information in memory, which is a well-understood, routine conventional activity. See MPEP § 2106.05(d)(II)(iv). Therefore, this does not amount to significantly more than the judicial exception.) Regarding claim 10 , the rejection of claim 9 is incorporated herein. Further, claim 10 recites the following additional element: wherein the feature information about each of the vehicles includes at least one of a location, a speed, a heading angle, a heading angle rate, or a driving lane of each of the vehicles, or any combination thereof. (This falls under mere instructions to apply an exception. See MPEP § 2106.05(f). Therefore, this does not amount to significantly more than the judicial exception.) Claims 11, 15 - 20 recite substantially similar subject matter to claims 1, 5 – 10 respectively and are rejected with the same rationale, mutatis mutandis . Claim Rejections - 35 USC § 103 07-06 AIA 15-10-15 In the event the determination of the status of the application as subject to AIA 35 U.S.C. 102 and 103 (or as subject to pre-AIA 35 U.S.C. 102 and 103) is incorrect, any correction of the statutory basis (i.e., changing from AIA to pre-AIA) for the rejection will not be considered a new ground of rejection if the prior art relied upon, and the rationale supporting the rejection, would be the same under either status. 07-20-aia AIA 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. 07-20-02-aia AIA This application currently names joint inventors. In considering patentability of the claims the examiner presumes that the subject matter of the various claims was commonly owned as of the effective filing date of the claimed invention(s) absent any evidence to the contrary. Applicant is advised of the obligation under 37 CFR 1.56 to point out the inventor and effective filing dates of each claim that was not commonly owned as of the effective filing date of the later invention in order for the examiner to consider the applicability of 35 U.S.C. 102(b)(2)(C) for any potential 35 U.S.C. 102(a)(2) prior art against the later invention. 07-21-aia AIA Claim s 1, 5 – 6, 9 – 11, 15 - 16, 19 – 20 are rejected under 35 U.S.C. 103 as being unpatentable over Elluswamy et al. (U.S. Pub. 2020/0249685 A1) in view of Taghavi (U.S. Pub. 2019/0050693 A1), further in view of Ramakrishnan et al. (U.S. Pub. 2021/0331695 A1) . Regarding claim 1 , Elluswamy teaches An apparatus for path prediction model training, the apparatus comprising: a sensor configured to obtain a time series of training images in a real environment; ([0011] of Elluswamy states “For example, sensors affixed to a vehicle capture data such as image data of the road and the surrounding environment a vehicle is driving on. The sensor data may capture vehicle lane lines, vehicle lanes, other vehicle traffic, obstacles, traffic control signs, etc. “ [0015] of Elluswamy states “In some embodiments, sensor data is received. The sensor data may include an image (such as video and/or still images), radar, audio, lidar, inertia, odometry, location, and/or other forms of sensor data. The sensor data includes a group of time series elements. For example, a group of time series elements may include a group of images captured from a camera sensor of a vehicle over a time period.” [0034] of Elluswamy states “At 301, elements of a time series are received. In various embodiments, the elements are sensor data such as image data captured at a vehicle and transmitted to a training server. The sensor data is captured over a period of time to create a time series of elements. In various embodiments, the elements are timestamps to maintain an ordering of the elements. As the elements progress through the time series, the events further in the time series are used to help predict an outcome from an earlier element of the time series.”) and a controller configured to train a path prediction model based on dynamic objects in the time series of training images, ([0011] of Elluswamy states “A machine learning training technique for generating highly accurate machine learning results is disclosed… The training data set is used to train a machine learning model for generating highly accurate machine learning results. In some embodiments, a time series of captured data is used to generate the training data.” [0013] of Elluswamy states “In various embodiments, the selected image and ground truth may apply to different features such as lane lines, path prediction for vehicles including neighboring vehicles, depth distances of objects, traffic control signs, etc. For example, a series of images of a vehicle in an adjacent lane is used to predict that vehicle's path. Using the time series of images and the actual path taken by the adjacent vehicle, a single image of the group and the actual path taken can be used as training data to predict the path of the vehicle.” [0019] of Elluswamy states “For example, one or more front-facing and/or pillar cameras capture lane markings of the lane the vehicle is traveling in. As another example, cameras capture neighboring vehicles including those attempting to cut into the lane the vehicle is traveling in. Additional sensors capture odometry, location, and/or vehicle control information including information related to vehicle trajectory.“ [0049] of Elluswamy states “In some embodiments, the various outputs of deep learning are used to construct a three-dimensional representation of the vehicle's environment for autonomous driving which includes predicted paths of vehicles, identified obstacles, identified traffic control signals including speed limits, etc. In some embodiments, the vehicle control module utilizes the determined results to control the vehicle along a determined path.”) train the path prediction model ([0013] of Elluswamy states “In various embodiments, the selected image and ground truth may apply to different features such as lane lines, path prediction for vehicles including neighboring vehicles, depth distances of objects, traffic control signs, etc.” [0015] of Elluswamy states “In some embodiments, a processor is used to train a machine learning model using the training dataset. For example, the training dataset is used to train a machine learning model for inferring features used for self-driving or driver-assisted operation of a vehicle. Using the trained machine learning model, a neural network can infer features associated with autonomous driving such as vehicle lanes, drivable space, objects (e.g., pedestrians, stationary vehicles, moving vehicles, etc.), weather (e.g., rain, hail, fog, etc.), traffic control objects (e.g., traffic lights, traffic signs, street signs, etc.), traffic patterns, etc.”) Elluswamy does not explicitly teach that generate a first table in which dynamic objects in a training image at a current time point are recorded, generate a second table in which dynamic objects in a training image at a future time point are recorded, assign, to a field of a first dynamic object in the second table, a token indicating that the first dynamic object is a dynamic object that disappears at the future time point when the first dynamic object in the training image at the current time point disappears from the training image at the future time point, and train … using the token. However, Ramakrishnan teaches generate a first table in which dynamic objects in a training image at a current time point are recorded, ([0002] of Ramakrishnan states “Specifically, the present disclosure relates to a system and method that applies machine learning techniques in a deep learning model to fuse data from multiple sensors to detect and identify objects, and observe and predict tracks of those objects in a field of view of autonomous or driverless vehicles” [0009] of Ramakrishnan states “Embodiments of the present disclosure accomplish this by applying input data collected from a camera to an agriculturally-focused deep learning model, and processes this input data by generating a list of objects identified in a current image frame of a field of view the camera, together with corresponding bounding boxes of those objects. Embodiments of the present disclosure then calculate the orientation of each object in the current image frame with respect to data collected by a ranging system such as radar, and correlates objects detected by the ranging system with the objects seen in the current image frame as seen by the camera using their orientations to fuse (e.g., associate, match, assemble or the like) the corresponding camera detections with detections from the ranging system.” Under BRI, a “table” is an organized data structure that records entries with associated fields. Ramakrishnan’s per-frame list of objects with associated bounding box data is exactly such a structure, where each row is a detected object and each column is a datafield. ) Also, Taghavi teaches generate a second table in which dynamic objects in a training image at a future time point are recorded, ([0048] of Taghavi states “It is assumed that frame k−1 310 has been fully annotated. In frame k−1 310, four objects are detected and assigned respective unique target IDs, specifically first detected object 330 a is assigned A, second detected object 332 a is assigned B, third detected object 334 a is assigned C and fourth detected object 336 is assigned D. Each detected object is also assigned an object label and track type (not shown). The annotations of frame k−1 310 are used for annotating frame k 320.” [0049] of Taghavi states ”In next frame k 320, four objects are detected. The four detected objects are indexed as object 1 332 b, object 2 330 b, object 3 334 b and object 4 338.” [0043] of Taghavi states “At the ninth frame 225, the object is no longer detected (indicated by dashed lines). Thus, for subsequent frames 230 starting from the ninth frame 225, the track is dead and annotation for the object is removed from those frames 230.”[0075] of Taghavi states “In the annotated dataset, each identified object in each frame has been labeled with temporal information, in addition to non-temporal object labels. Annotation with temporal information includes target IDs and track types that have been assigned to each identified object in each frame. As noted previously, such information is considered to be temporal because they take into account changes (e.g., motion) of an object through two or more frames.” Taghavi generates a per-frame annotation record for frame k containing each detected object with its assigned target ID, object label, and track type. Under BRI as described above, this per-frame annotation record, where each object is a row and each assigned field is a column could be the claimed second table.) assign, to a field of a first dynamic object in the second table, a token indicating that the first dynamic object is a dynamic object that disappears at the future time point when the first dynamic object in the training image at the current time point disappears from the training image at the future time point, ([0069] of Taghavi states “At 412, if an object in frame k cannot be assigned an existing target ID (e.g., all existing target IDs have been assigned to other objects in the frame, or the likelihood calculated for all existing target IDs is below the predetermined threshold), a new target ID may be initiated and assigned to that object. This indicates that the object associated with the new target ID is newly appearing in frame k.” [0070] of Taghavi states “It should be noted that an existing target ID may be unassigned to any object in frame k, indicating that the object associated with that target ID has not been found in frame k.” [0042] of Taghavi states “A confirmed track is demoted to dead track when the associated object no longer satisfies the requirements for confirmed track, as defined by the track management scheme… At the ninth frame 225, the object is no longer detected (indicated by dashed lines). Thus, for subsequent frames 230 starting from the ninth frame 225, the track is dead and annotation for the object is removed from those frames 230.” For a disappearing object, Taghavi does two things in the frame k annotation record. IT leaves the target ID unassigned or assigns dead to the track type field. [0005] of Taghavi states “The method includes iterating the calculating, assigning and performing for all objects in all frames of the dataset, and outputting the annotated dataset, wherein the annotated dataset contains target IDs and track types assigned to all objects in all frames.” [0037] of Taghavi states “A track management scheme defines how a track is identified as one of several track types, for example tentative, confirmed or dead. By annotating the track type of an object over a sequence of frames, information about the motion of the tracked object provided in the dataset.” [0048] of Taghavi states “It is assumed that frame k−1 310 has been fully annotated. In frame k−1 310, four objects are detected and assigned respective unique target IDs, specifically first detected object 330 a is assigned A, second detected object 332 a is assigned B, third detected object 334 a is assigned C and fourth detected object 336 is assigned D. Each detected object is also assigned an object label and track type (not shown). The annotations of frame k−1 310 are used for annotating frame k 320.” [0049] of Taghavi states ”In next frame k 320, four objects are detected. The four detected objects are indexed as object 1 332 b, object 2 330 b, object 3 334 b and object 4 338. It should be noted that the objects 330 b, 332 b, 334 b, 338 detected in frame k 320 are not necessarily indexed in the same order as the objects 330 a, 332 a, 334 a, 336 detected in frame k−1 310.” Taghavi teaches annotating objects across consecutive frames using target IDs and track types. Frame k-1 corresponds to the earlier/current time point, and frame k corresponds to the later/future time point. Taghavi further teaches that when an existing target ID is not assigned to any object in frame k, this indicates that the object associated with that target ID “has not been found in frame k.” Under BRI, the claimed “field” reads on an annotation entry for target-id, track type, or object status information, and the claimed “token” reads on a value or status indicating that the object is not found, no longer detected, or dead at the frame. Taghavi’s explanation about dead track teaches that a confirmed track may be demoted to dead when the associated object is no longer detected. Therefore, Taghavi teaches assigning an object status token indicating that a dynamic object present at the current time point disappears or is no longer detected at the future time point. ) and train … using the token. ([0004] of Taghavi states “The disclosed examples enable different track management schemes to be automatically applied for annotating a dataset. The resulting annotated dataset may be used for training DNN tracking algorithms.” [0019] of Taghavi states “The temporal information includes target identifiers (IDs) and track type assigned to each object in each frame of the training dataset. The method also includes adjusting weighting of the deep tracking neural network to reduce the first error vector.” POSITA would have been motivated to incorporate Taghavi’s temporal disappearance annotations into Elluswamy’s path prediction training data because accounting for whether dynamic objects remain or disappear between frames directly improves path prediction accuracy in real environment.) It would have been obvious to one with ordinary skill in the art before the effective filing date of the invention to combine the teachings of Elluswamy, Ramakrishnan, and Taghavi. Elluswamy teaches using timeseries sensor images data of vehicles in real driving environments to train a model that predicts vehicle paths and lane behavior. Ramakrishnan teaches representing dynamic object by a feature vector as input to prediction logic. Taghavi teaches assigning object IDs and track type labels (e.g., tentative, confirmed, dead, etc.) for objects appearing across different time frames and using those labels to train a network. One of ordinary skill in the art would have been motivated to incorporate the teachings of Taghavi, Ramakrishan into Elluswamy’s path predicting training to accurately handle objects entering and leaving the field of view. It would have been predictable combination for the invention regarding vehicles to obtain a more robust path prediction model using time series images dealing with newly appearing and disappearing objects. Regarding claim 5 , the rejection of claim 2 is incorporated herein. Furthermore, the combination of Elluswamy, Taghavi, and Ramakrishnan teach wherein the controller is configured to train a second dynamic object that is not present in a training image at a past time point as a dynamic object newly appearing in the training image at the current time point when the second dynamic object newly appears in the training image at the current time point. ([0005] of Taghavi states “For an object of interest in a frame of interest, within a plurality of frames of a dataset, the method includes calculating one or more likelihood functions to correlate feature score of the object of interest with respective feature scores each associated with one or more previously assigned target identifiers (IDs) in a selected range of frames.” [0012] of Taghavi states “The instructions further cause the system to assign a target ID to the object of interest by: identifying a previously assigned target ID associated with a calculated highest likelihood and assigning the identified target ID to the object of interest; or initiating a new target ID when none of the previously assigned target IDs is associated with a calculated likelihood that satisfies a predetermined threshold and assigning the initiated new target ID to the object of interest.” If an object has no matching previously assigned ID (meaning it didn’t appear in past frames), it will be given new target ID. It is the second dynamic object that newly appeared to be used.) Regarding claim 6 , the rejection of claim 5 is incorporated herein. Furthermore, the combination of Elluswamy, Taghavi, and Ramakrishnan teach wherein the controller is configured to determine a value indicating that the second dynamic object is the dynamic object newly appearing at the current time point and train the path prediction model using the value. ([0005] of Taghavi states “For an object of interest in a frame of interest, within a plurality of frames of a dataset, the method includes calculating one or more likelihood functions to correlate feature score of the object of interest with respective feature scores each associated with one or more previously assigned target identifiers (IDs) in a selected range of frames.” [0012] of Taghavi states “The instructions further cause the system to assign a target ID to the object of interest by: identifying a previously assigned target ID associated with a calculated highest likelihood and assigning the identified target ID to the object of interest; or initiating a new target ID when none of the previously assigned target IDs is associated with a calculated likelihood that satisfies a predetermined threshold and assigning the initiated new target ID to the object of interest.” [0091] of Taghavi states “FIG. 7 illustrates an example DNN 700 that may be trained using a dataset that has been annotated with temporal information, such as the annotated dataset outputted by the method 400. The DBB 700 may be trained, such as using the method 500, to generate a deep tracking neural network for object tracking.”) Regarding claim 9 , the rejection of claim 1 is incorporated herein. Furthermore, the combination of Elluswamy, Taghavi, and Ramakrishnan teach the dynamic objects in the time series of training images are vehicles, and the controller is configured to extract feature information about each of the vehicles in the time series of training images as an input vector for the path prediction model. ([0023] of Ramakrishnan states “The object detection and tracking framework 100 of the present disclosure therefore contemplates that input data 110 may include first vehicular state data 120 and second vehicular state data 121, each of which comprise multiple operational characteristics of an autonomously-operated vehicle 102. It is to be understood however, and as noted above, that any number of autonomously-operated vehicles 102 may be deployed at least in the performance of agricultural activities 108 (and for transportation-related activities, for example where an unmanned service vehicle is deployed to assist a stranded passenger or commercial vehicle), and therefore the vehicular state data may include operational characteristics for n autonomously-operated vehicles 102, and embodiments of the present disclosure are not to be limited to any specific number of such autonomously-operated vehicles 102, either for the vehicular state data or for any other aspect of the present disclosure.” [0024] of Ramakrishnan states “These operational characteristics may include a latitude 122 and a longitude 123, representing positional coordinates of the respective autonomously-operated vehicle 102.” [0025] of Ramakrishnan states “The first vehicular state data 120 and the second vehicular state data 121 may also include operational characteristics associated with vehicular movement, such as speed 124, heading 125, yaw-rate 126, and curvature 127.”) Regarding claim 10 , the rejection of claim 9 is incorporated herein. Furthermore, the combination of Elluswamy, Taghavi, and Ramakrishnan teach wherein the feature information about each of the vehicles includes at least one of a location, a speed, a heading angle, a heading angle rate, or a driving lane of each of the vehicles, or any combination thereof. ([0011] of Elluswamy states “The sensor data may capture vehicle lane lines, vehicle lanes, other vehicle traffic, obstacles, traffic control signs, etc. Odometry and other similar sensors capture vehicle operating parameters such as vehicle speed, steering, orientation, change in direction, change in location, change in elevation, change in speed, etc.”) Claims 11, 15 – 16, 19 – 20 recite substantially similar subject matter as claims 1, 5 – 6, 9 – 10 respectively, and are rejected with the same rationale, mutatis mutandis . 07-21-aia AIA Claim s 7 – 8, 17 – 18 are rejected under 35 U.S.C. 103 as being unpatentable over Elluswamy et al. (U.S. Pub. 2020/0249685 A1) in view of Taghavi (U.S. Pub. 2019/0050693 A1), Ramakrishnan et al. (U.S. Pub. 2021/0331695 A1), further in view of Wang et al. (U.S. Pub. 2023/0211660 A1) . Regarding claim 7 , the rejection of claim 1 is incorporated herein. Furthermore, the combination of Elluswamy, Taghavi, and Ramakrishnan does not teach wherein the path prediction model is a transformer network. However, Wang teaches that wherein the path prediction model is a transformer network. ([0024] of Wang states “After sufficient data associated with the driver is collected, the data may be input to a transformer network as training data and the transformer network may be trained based on the training data to predict the vehicle acceleration at a future time step based on the data associated with a current time step.” [0079] of Wang states “When sufficient training data is received by the server, the server may train a transformer network to predict future vehicle trajectories, which indicates predicted driving behavior of the driver based on current driving conditions.”) It would have been obvious to one with ordinary skill in the art before the effective filing date of the invention to combine the teachings of Wang with the combination of Elluswamy, Taghavi, and Ramakrishnan. Elluswamy teaches using timeseries sensor images data of vehicles in real driving environments to train a model that predicts vehicle paths and lane behavior. Ramakrishnan teaches representing dynamic object by a feature vector as input to prediction logic. Taghavi teaches assigning object IDs and track type labels (e.g., newly appearing or dead) for objects appearing across different time frames and using those labels to train a network. Wang further teaches implementing vehicle behavior and trajectory prediction using transformer that input vectorized driving data to predict future vehicle trajectories. One of ordinary skill in the art would have been motivated to incorporate the teachings of Wang into the combination of Elluswamy, Ramakrishnan, and Taghavi to improve the system with transformer architectures because transformer is a well-known technique for models dealing with sequential data and Wang is using it to apply on vehicle trajectory prediction. It would have been predictable combination to implement the path prediction model with transformer architecture for more robust and accurate training model. Regarding claim 8 , the rejection of claim 7 is incorporated herein. Furthermore, the combination of Elluswamy, Taghavi, Ramakrishnan, and Wang teaches wherein the transformer network is configured to train a location of each of the dynamic objects at a future time point based on an input vector of each of the dynamic objects at a past time point and an input vector of each of the dynamic objects at a current time point. ([0024] of Wang states “After sufficient data associated with the driver is collected, the data may be input to a transformer network as training data and the transformer network may be trained based on the training data to predict the vehicle acceleration at a future time step based on the data associated with a current time step.” [0062] of Wang states “Referring back to FIG. 3 , the transformer training module 322 may train the transformer network 600 to predict future trajectories of road agents based on training data comprising past driving data collected by the ego vehicle 104.” [0079] of Wang states “When sufficient training data is received by the server, the server may train a transformer network to predict future vehicle trajectories, which indicates predicted driving behavior of the driver based on current driving conditions.”) Claims 17 – 18 recite substantially similar subject matter as claims 7 – 8 respectively, and are rejected with the same rationale, mutatis mutandis. Conclusion 07-40 AIA 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 BYUNGKWON HAN whose telephone number is (571)272-5294. The examiner can normally be reached M-F: 9:00AM-6PM PST. 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, Li B Zhen can be reached at (571)272-3768. 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. /BYUNGKWON HAN/ Examiner, Art Unit 2121 /Li B. Zhen/Supervisory Patent Examiner, Art Unit 2121 Application/Control Number: 18/136,687 Page 2 Art Unit: 2121 Application/Control Number: 18/136,687 Page 3 Art Unit: 2121 Application/Control Number: 18/136,687 Page 4 Art Unit: 2121 Application/Control Number: 18/136,687 Page 5 Art Unit: 2121 Application/Control Number: 18/136,687 Page 6 Art Unit: 2121 Application/Control Number: 18/136,687 Page 7 Art Unit: 2121 Application/Control Number: 18/136,687 Page 8 Art Unit: 2121 Application/Control Number: 18/136,687 Page 9 Art Unit: 2121 Application/Control Number: 18/136,687 Page 10 Art Unit: 2121 Application/Control Number: 18/136,687 Page 11 Art Unit: 2121 Application/Control Number: 18/136,687 Page 12 Art Unit: 2121 Application/Control Number: 18/136,687 Page 13 Art Unit: 2121 Application/Control Number: 18/136,687 Page 14 Art Unit: 2121 Application/Control Number: 18/136,687 Page 15 Art Unit: 2121 Application/Control Number: 18/136,687 Page 16 Art Unit: 2121 Application/Control Number: 18/136,687 Page 17 Art Unit: 2121 Application/Control Number: 18/136,687 Page 18 Art Unit: 2121 Application/Control Number: 18/136,687 Page 19 Art Unit: 2121 Application/Control Number: 18/136,687 Page 20 Art Unit: 2121 Application/Control Number: 18/136,687 Page 21 Art Unit: 2121 Application/Control Number: 18/136,687 Page 22 Art Unit: 2121
Prosecution Insights