DETAILED ACTION
This office action is in response to the amendment filed on 3/18/2026. In the amendment, claims 1-2, 21-22 and 29-30 have been amended. Overall, claims 1-30 are pending in this application.
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 .
Claim Interpretation
The following is a quotation of 35 U.S.C. 112(f):
(f) Element in Claim for a Combination. – An element in a claim for a combination may be expressed as a means or step for performing a specified function without the recital of structure, material, or acts in support thereof, and such claim shall be construed to cover the corresponding structure, material, or acts described in the specification and equivalents thereof.
The following is a quotation of pre-AIA 35 U.S.C. 112, sixth paragraph:
An element in a claim for a combination may be expressed as a means or step for performing a specified function without the recital of structure, material, or acts in support thereof, and such claim shall be construed to cover the corresponding structure, material, or acts described in the specification and equivalents thereof.
The claims in this application are given their broadest reasonable interpretation using the plain meaning of the claim language in light of the specification as it would be understood by one of ordinary skill in the art. The broadest reasonable interpretation of a claim element (also commonly referred to as a claim limitation) is limited by the description in the specification when 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, is invoked.
As explained in MPEP § 2181, subsection I, claim limitations that meet the following three-prong test will be interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph:
(A) the claim limitation uses the term “means” or “step” or a term used as a substitute for “means” that is a generic placeholder (also called a nonce term or a non-structural term having no specific structural meaning) for performing the claimed function;
(B) the term “means” or “step” or the generic placeholder is modified by functional language, typically, but not always linked by the transition word “for” (e.g., “means for”) or another linking word or phrase, such as “configured to” or “so that”; and
(C) the term “means” or “step” or the generic placeholder is not modified by sufficient structure, material, or acts for performing the claimed function.
Use of the word “means” (or “step”) in a claim with functional language creates a rebuttable presumption that the claim limitation is to be treated in accordance with 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph. The presumption that the claim limitation is interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, is rebutted when the claim limitation recites sufficient structure, material, or acts to entirely perform the recited function.
Absence of the word “means” (or “step”) in a claim creates a rebuttable presumption that the claim limitation is not to be treated in accordance with 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph. The presumption that the claim limitation is not interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, is rebutted when the claim limitation recites function without reciting sufficient structure, material or acts to entirely perform the recited function.
Claim limitations in this application that use the word “means” (or “step”) are being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, except as otherwise indicated in an Office action. Conversely, claim limitations in this application that do not use the word “means” (or “step”) are not being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, except as otherwise indicated in an Office action.
A review of the specification shows that the following appears to be the corresponding structure described in the specification for the 35 U.S.C. 112(f) limitation:
means for applying (Corresponding structure that performs the recited function in claim 29 is control program or logic with processor).
means for performing (Corresponding structure that performs the recited function in claim 29 is control program or logic with processor).
Claim Rejections - 35 USC § 103
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.
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.
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.
Claim(s) 1-3, 5-6, 8, 15, 17-23 and 25-30 are rejected under 35 U.S.C. 103 as being unpatentable over Pub No. US 2023/0399023 A1 to Li et. al. (Li) in view of Pub No. US 2020/0207339 A1 to Neil et. al. (Neil).
In Reference to Claim 1
Li teaches (except for the bolded and italic recitations below):
An ego vehicle (100), comprising:
one or more memories (114) (see at least Li Figs. 1-2 and paragraph [0080]);
one or more transceivers (123) (see at least Li Figs. 1-2 and paragraph [0094]); and
one or more processors (113) communicatively coupled to the one or more memories (114) and the one or more transceivers (123), the one or more processors (113), either alone or in combination (see at least Li Figs. 1-2 and paragraph [0080]-[0081], [0094]-[0095]), configured to:
apply a machine learning model (30) (see at least Li Fig. 3 and paragraph [0106]) to one or more agent tensors (Li teaches at least in paragraph [0160] “the interaction feature between the target vehicle and each of the other vehicles is input into an interaction feature vector extraction network corresponding to the target vehicle and the other vehicle, to extract an interaction feature vector between the target vehicle and the other vehicle. The vector represents impact of the other vehicle on the driving intention of the target vehicle”) and one or more map tensors (Li teaches at least in paragraph [0171] “The interaction feature between the target vehicle and the road associated with each lane is input into the first road feature extraction subnetwork, to extract an interaction feature vector between the target vehicle and the road associated with each lane; and the interaction feature vector between the target vehicle and the road associated with each lane is input into the second road feature extraction subnetwork, to extract an interaction feature implicit vector between the target vehicle and the road associated with each lane”) associated with a target vehicle to obtain a predicted drive intention of the target vehicle at a roadway intersection (see Li Figs. 6-8), wherein the predicted drive intention comprises a turn classification, a flow classification, or both (see at least Li Fig. 3 and paragraph [0106]), and wherein the one or more agent tensors are three-dimensional tensors (Li teaches at least in paragraphs [0149], [0150] and [0151] “a method for obtaining the interaction feature between the target vehicle and the another vehicle may be extracted according to the following rule: extracting one or more of a location feature of each of the surrounding vehicles in a first coordinate system, a speed feature, and a head orientation feature, where an origin of the first coordinate system is a current location of the target vehicle, the first coordinate system is a rectangular coordinate system, a y-axis of the first coordinate system is parallel to a length direction of a vehicle body of the target vehicle, and a forward direction of the y-axis is consistent with a head orientation of the target vehicle”, “In step S503, a driving feature of the target vehicle relative to each lane, namely, an interaction feature between the target vehicle and each lane, is determined based on the driving information of the target vehicle and the lane layer information” and “the interaction feature between the target vehicle and each lane may be extracted according to the following rule: extracting one or more of a location feature of a target vehicle in each third coordinate system, a feature of an angle formed by the head orientation of the target vehicle and a driving direction of the lane, and a feature that a location of the target vehicle in each third coordinate system, and the angle formed by the head orientation of the target vehicle and the driving orientation of the lane change with the driving moment, where each third coordinate system is a frenet coordinate system, a reference line of each third coordinate system is determined based on a center line of each lane, and an origin of each third coordinate system is determined based on an end point of the center line of each lane”); and
perform a driving maneuver based on the predicted drive intention (Li teaches at least in paragraph [0122] “The prediction unit 43 predicts the behavior intention and the future track of the target vehicle based on current map information and the target information sensed by the sensing unit. The planning unit 44 plans a driving route of the vehicle based on a prediction result of the prediction unit and/or output information of the navigation unit 47. The control unit 45 controls, based on the driving route planned by the planning unit, the vehicle to drive on the planned driving route. The target fusion unit 42, the prediction unit 43, the planning unit 43, and the control unit 45 are all implemented in the processor in FIG. 1 or FIG. 2. In a driving process of the vehicle, an intention of another vehicle is predicted in real time, accurately, and reliably, so that the vehicle can predict a traffic condition in front of the vehicle, and establish a traffic situation around the vehicle”).
Li teaches to obtain a predicted drive intention of the target vehicle at a roadway intersection however Li does not explicitly teach (bolded and italic recitations above) as to the predicted drive intention comprises a turn classification, a flow classification, or both (which are maneuver model of the vehicles). However, it is known in the art before the effective filing date of the claimed invention that to determine the predicted drive intention comprises a turn classification, a flow classification, or both. For example, Neil teaches to predict a maneuver that an object in a driving environment of the autonomous vehicle and to determine the predicted drive intention comprises a turn classification (label), a flow classification (label), or both. Neil further implicitly teaches that having such classifications (labels) provides efficiency of generates predictions of maneuvers that are to be undertaken by objects in environments of autonomous vehicles (see at least Neil Figs. 1-2 and 8-9 and paragraphs [0008]-[0009] and [0034]-[0035]). Therefore it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify the system of Li to include the turn classification, a flow classification, or both in predicted drive intention as taught by Neil in order to provide efficiency of generates predictions of maneuvers that are to be undertaken by objects in environments of autonomous vehicles.
In Reference to Claim 2
The ego vehicle of claim 1 (see rejection to claim 1 above), wherein the one or more agent tensors representing:
a plurality of polylines representing trajectories of the target vehicle and one or more neighbor vehicles of the target vehicle over a most recent period of time, a plurality of points of each of the plurality of polylines, and a plurality of features of each of the plurality of polylines (Li teaches at least in paragraph [0149] “a method for obtaining the interaction feature between the target vehicle and the another vehicle may be extracted according to the following rule: extracting one or more of a location feature of each of the surrounding vehicles in a first coordinate system, a speed feature, and a head orientation feature, where an origin of the first coordinate system is a current location of the target vehicle, the first coordinate system is a rectangular coordinate system, a y-axis of the first coordinate system is parallel to a length direction of a vehicle body of the target vehicle, and a forward direction of the y-axis is consistent with a head orientation of the target vehicle”).
In Reference to Claim 3
The ego vehicle of claim 2 (see rejection to claim 2 above), wherein the plurality of features comprises: x coordinates of the target vehicle and the one or more neighbor vehicles at each of the plurality of points, y coordinates of the target vehicle and the one or more neighbor vehicles at each of the plurality of points, previous x coordinates of the target vehicle and the one or more neighbor vehicles at each of the plurality of points, previous y coordinates of the target vehicle and the one or more neighbor vehicles at each of the plurality of points, x-axis velocity values of the target vehicle and the one or more neighbor vehicles at each of the plurality of points, y-axis velocity values of the target vehicle and the one or more neighbor vehicles at each of the plurality of points, angular velocity values of the target vehicle and the one or more neighbor vehicles at each of the plurality of points, acceleration values of the target vehicle and the one or more neighbor vehicles at each of the plurality of points, blinker states of the target vehicle and the one or more neighbor vehicles at each of the plurality of points, a time offset between a current timestamp of the ego vehicle and a timestamp of each of the plurality of points, mask values of the target vehicle and the one or more neighbor vehicles at each of the plurality of points, or any combination thereof (Li teaches at least in paragraph [0149] “a method for obtaining the interaction feature between the target vehicle and the another vehicle may be extracted according to the following rule: extracting one or more of a location feature of each of the surrounding vehicles in a first coordinate system, a speed feature, and a head orientation feature, where an origin of the first coordinate system is a current location of the target vehicle, the first coordinate system is a rectangular coordinate system, a y-axis of the first coordinate system is parallel to a length direction of a vehicle body of the target vehicle, and a forward direction of the y-axis is consistent with a head orientation of the target vehicle”).
In Reference to Claim 5
The ego vehicle of claim 1 (see rejection to claim 1 above), wherein the one or more map tensors are three-dimensional tensors representing: a plurality of polylines representing lanes, lane boundaries, or both along which the target vehicle and one or more neighbor vehicles are traveling, a plurality of points of each of the plurality of polylines, and a plurality of features of each of the plurality of polylines (Li teaches at least in paragraph [0151] “the interaction feature between the target vehicle and each lane may be extracted according to the following rule: extracting one or more of a location feature of a target vehicle in each third coordinate system, a feature of an angle formed by the head orientation of the target vehicle and a driving direction of the lane, and a feature that a location of the target vehicle in each third coordinate system, and the angle formed by the head orientation of the target vehicle and the driving orientation of the lane change with the driving moment, where each third coordinate system is a frenet coordinate system, a reference line of each third coordinate system is determined based on a center line of each lane, and an origin of each third coordinate system is determined based on an end point of the center line of each lane”).
In Reference to Claim 6
The ego vehicle of claim 5 (see rejection to claim 5 above), wherein the plurality of features comprises: x coordinates of the lanes, lane boundaries, or both along which the target vehicle and the one or more neighbor vehicles are traveling, y coordinates of the lanes, lane boundaries, or both along which the target vehicle and the one or more neighbor vehicles are traveling, x directions of the lanes, lane boundaries, or both along which the target vehicle and the one or more neighbor vehicles are traveling, y directions of the lanes, lane boundaries, or both along which the target vehicle and the one or more neighbor vehicles are traveling, previous x coordinates of the lanes, lane boundaries, or both along which the target vehicle and the one or more neighbor vehicles are traveling, previous y coordinates of the lanes, lane boundaries, or both along which the target vehicle and the one or more neighbor vehicles are traveling, point types of the lanes, lane boundaries, or both along which the target vehicle and the one or more neighbor vehicles are traveling, mask values of the target vehicle at each of the plurality of points, or any combination thereof (Li teaches at least in paragraph [0151] “the interaction feature between the target vehicle and each lane may be extracted according to the following rule: extracting one or more of a location feature of a target vehicle in each third coordinate system, a feature of an angle formed by the head orientation of the target vehicle and a driving direction of the lane, and a feature that a location of the target vehicle in each third coordinate system, and the angle formed by the head orientation of the target vehicle and the driving orientation of the lane change with the driving moment, where each third coordinate system is a frenet coordinate system, a reference line of each third coordinate system is determined based on a center line of each lane, and an origin of each third coordinate system is determined based on an end point of the center line of each lane”).
In Reference to Claim 8
The ego vehicle of claim 5 (see rejection to claim 5 above), wherein the point types of the lanes, lane boundaries, or both along which the target vehicle and the one or more neighbor vehicles are traveling comprise: center point types, boundary point types, or a combination thereof (Li teaches at least in Fig.7 and paragraph [0152] “It is easy to understand that a center line of a lane is a line formed by sequentially connecting center points in a width direction of the lane from a start point of the lane to an end point of the lane”).
In Reference to Claim 15
The ego vehicle of claim 1 (see rejection to claim 1 above), wherein the turn classification represents a probability that the target vehicle will perform one of a plurality of classes of turns (Neil teaches at least in Figs.2 and 9 and paragraph [0008], [0034] and [0062] “the maneuver label is indicative of a maneuver that the object executes during the time period. For instance, the maneuver label may be straight, left lane change, right lane change, left turn, right turn, stationary, or unknown” “As shown in FIG. 2, the maneuver labels 202-214 include a straight label 202, a left lane change label 204, a right lane change label 206, a left turn label 208, a right turn label 210, a stationary label 212, and an unknown label 214. The unknown label 214 captures maneuvers executed by objects that are not captured by the maneuver labels 202-212. In an example, the maneuver label generation application 106 can assign one of the maneuver labels 202-214 to the sensor data 1 216 and one of the maneuver labels 202-214 to the sensor data M 218. Thus, the maneuver label generation application 106 generates labeled sensor data” and “As such, it may be ambiguous as to whether the vehicle 902 is to execute a left hand turn 908, continue a straight heading 910, or execute a (wide) right turn 912. Using the above-described processes, the autonomous vehicle 400 may provide sensor data generated by the sensor systems 402-404 as input to the machine learning model 502. The machine learning model 502 may then output an indication of a maneuver that the vehicle 902 is predicted to execute. For instance, the machine learning model 502 may be configured to output a probability distribution over the left hand turn 908, the straight heading 910, and the right turn 912. The autonomous vehicle 400 may then operate based upon the indication of the maneuver that the vehicle 902 is predicted to execute”).
In Reference to Claim 17
The ego vehicle of claim 1 (see rejection to claim 1 above), wherein the flow classification represents a probability that the target vehicle will perform one of a plurality of classes of flows (Neil teaches at least in Figs.2 and 9 and paragraph [0008], [0034] and [0062] “the maneuver label is indicative of a maneuver that the object executes during the time period. For instance, the maneuver label may be straight, left lane change, right lane change, left turn, right turn, stationary, or unknown” “As shown in FIG. 2, the maneuver labels 202-214 include a straight label 202, a left lane change label 204, a right lane change label 206, a left turn label 208, a right turn label 210, a stationary label 212, and an unknown label 214. The unknown label 214 captures maneuvers executed by objects that are not captured by the maneuver labels 202-212. In an example, the maneuver label generation application 106 can assign one of the maneuver labels 202-214 to the sensor data 1 216 and one of the maneuver labels 202-214 to the sensor data M 218. Thus, the maneuver label generation application 106 generates labeled sensor data” and “As such, it may be ambiguous as to whether the vehicle 902 is to execute a left hand turn 908, continue a straight heading 910, or execute a (wide) right turn 912. Using the above-described processes, the autonomous vehicle 400 may provide sensor data generated by the sensor systems 402-404 as input to the machine learning model 502. The machine learning model 502 may then output an indication of a maneuver that the vehicle 902 is predicted to execute. For instance, the machine learning model 502 may be configured to output a probability distribution over the left hand turn 908, the straight heading 910, and the right turn 912. The autonomous vehicle 400 may then operate based upon the indication of the maneuver that the vehicle 902 is predicted to execute”).
In Reference to Claim 18
The ego vehicle of claim 17 (see rejection to claim 17 above), wherein the plurality of classes of flows comprises: free flow, starting, stopping, and stopped (Neil teaches at least in Figs.2 and 9 and paragraph [0008], [0034] and [0062] “the maneuver label is indicative of a maneuver that the object executes during the time period. For instance, the maneuver label may be straight, left lane change, right lane change, left turn, right turn, stationary, or unknown” “As shown in FIG. 2, the maneuver labels 202-214 include a straight label 202, a left lane change label 204, a right lane change label 206, a left turn label 208, a right turn label 210, a stationary label 212, and an unknown label 214. The unknown label 214 captures maneuvers executed by objects that are not captured by the maneuver labels 202-212. In an example, the maneuver label generation application 106 can assign one of the maneuver labels 202-214 to the sensor data 1 216 and one of the maneuver labels 202-214 to the sensor data M 218. Thus, the maneuver label generation application 106 generates labeled sensor data” and “As such, it may be ambiguous as to whether the vehicle 902 is to execute a left hand turn 908, continue a straight heading 910, or execute a (wide) right turn 912. Using the above-described processes, the autonomous vehicle 400 may provide sensor data generated by the sensor systems 402-404 as input to the machine learning model 502. The machine learning model 502 may then output an indication of a maneuver that the vehicle 902 is predicted to execute. For instance, the machine learning model 502 may be configured to output a probability distribution over the left hand turn 908, the straight heading 910, and the right turn 912. The autonomous vehicle 400 may then operate based upon the indication of the maneuver that the vehicle 902 is predicted to execute”).
In Reference to Claim 19
The ego vehicle of claim 1 (see rejection to claim 1 above), wherein the predicted drive intention comprises a turn trajectory associated with the turn classification (Neil teaches at least in Figs.2 and 9 and paragraph [0008], [0034] and [0062] “the maneuver label is indicative of a maneuver that the object executes during the time period. For instance, the maneuver label may be straight, left lane change, right lane change, left turn, right turn, stationary, or unknown” “As shown in FIG. 2, the maneuver labels 202-214 include a straight label 202, a left lane change label 204, a right lane change label 206, a left turn label 208, a right turn label 210, a stationary label 212, and an unknown label 214. The unknown label 214 captures maneuvers executed by objects that are not captured by the maneuver labels 202-212. In an example, the maneuver label generation application 106 can assign one of the maneuver labels 202-214 to the sensor data 1 216 and one of the maneuver labels 202-214 to the sensor data M 218. Thus, the maneuver label generation application 106 generates labeled sensor data” and “As such, it may be ambiguous as to whether the vehicle 902 is to execute a left hand turn 908, continue a straight heading 910, or execute a (wide) right turn 912. Using the above-described processes, the autonomous vehicle 400 may provide sensor data generated by the sensor systems 402-404 as input to the machine learning model 502. The machine learning model 502 may then output an indication of a maneuver that the vehicle 902 is predicted to execute. For instance, the machine learning model 502 may be configured to output a probability distribution over the left hand turn 908, the straight heading 910, and the right turn 912. The autonomous vehicle 400 may then operate based upon the indication of the maneuver that the vehicle 902 is predicted to execute”).
In Reference to Claim 20
The ego vehicle of claim 1 (see rejection to claim 1 above), wherein the driving maneuver comprises: a lane change before or after the roadway intersection, a left turn at the roadway intersection, a right turn at the roadway intersection, a U-turn at the roadway intersection, driving straight through the roadway intersection, a merge into a lane on which the target vehicle is driving, or a hard braking event (Neil teaches at least in Figs.2 and 9 and paragraph [0008], [0034] and [0062] “the maneuver label is indicative of a maneuver that the object executes during the time period. For instance, the maneuver label may be straight, left lane change, right lane change, left turn, right turn, stationary, or unknown” “As shown in FIG. 2, the maneuver labels 202-214 include a straight label 202, a left lane change label 204, a right lane change label 206, a left turn label 208, a right turn label 210, a stationary label 212, and an unknown label 214. The unknown label 214 captures maneuvers executed by objects that are not captured by the maneuver labels 202-212. In an example, the maneuver label generation application 106 can assign one of the maneuver labels 202-214 to the sensor data 1 216 and one of the maneuver labels 202-214 to the sensor data M 218. Thus, the maneuver label generation application 106 generates labeled sensor data” and “As such, it may be ambiguous as to whether the vehicle 902 is to execute a left hand turn 908, continue a straight heading 910, or execute a (wide) right turn 912. Using the above-described processes, the autonomous vehicle 400 may provide sensor data generated by the sensor systems 402-404 as input to the machine learning model 502. The machine learning model 502 may then output an indication of a maneuver that the vehicle 902 is predicted to execute. For instance, the machine learning model 502 may be configured to output a probability distribution over the left hand turn 908, the straight heading 910, and the right turn 912. The autonomous vehicle 400 may then operate based upon the indication of the maneuver that the vehicle 902 is predicted to execute”).
In Reference to Claim 21
Li teaches (except for the bolded and italic recitations below):
A method of drive trajectory prediction performed by an ego vehicle, comprising:
applying a machine learning model (30) (see at least Li Fig. 3 and paragraph [0106]) to one or more agent tensors (Li teaches at least in paragraph [0160] “the interaction feature between the target vehicle and each of the other vehicles is input into an interaction feature vector extraction network corresponding to the target vehicle and the other vehicle, to extract an interaction feature vector between the target vehicle and the other vehicle. The vector represents impact of the other vehicle on the driving intention of the target vehicle”) and one or more map tensors (Li teaches at least in paragraph [0171] “The interaction feature between the target vehicle and the road associated with each lane is input into the first road feature extraction subnetwork, to extract an interaction feature vector between the target vehicle and the road associated with each lane; and the interaction feature vector between the target vehicle and the road associated with each lane is input into the second road feature extraction subnetwork, to extract an interaction feature implicit vector between the target vehicle and the road associated with each lane”) associated with a target vehicle to obtain a predicted drive intention of the target vehicle at a roadway intersection (see Li Figs. 6-8), wherein the predicted drive intention comprises a turn classification, a flow classification, or both (see at least Li Fig. 3 and paragraph [0106]), and wherein the one or more agent tensors are three-dimensional tensors (Li teaches at least in paragraphs [0149], [0150] and [0151] “a method for obtaining the interaction feature between the target vehicle and the another vehicle may be extracted according to the following rule: extracting one or more of a location feature of each of the surrounding vehicles in a first coordinate system, a speed feature, and a head orientation feature, where an origin of the first coordinate system is a current location of the target vehicle, the first coordinate system is a rectangular coordinate system, a y-axis of the first coordinate system is parallel to a length direction of a vehicle body of the target vehicle, and a forward direction of the y-axis is consistent with a head orientation of the target vehicle”, “In step S503, a driving feature of the target vehicle relative to each lane, namely, an interaction feature between the target vehicle and each lane, is determined based on the driving information of the target vehicle and the lane layer information” and “the interaction feature between the target vehicle and each lane may be extracted according to the following rule: extracting one or more of a location feature of a target vehicle in each third coordinate system, a feature of an angle formed by the head orientation of the target vehicle and a driving direction of the lane, and a feature that a location of the target vehicle in each third coordinate system, and the angle formed by the head orientation of the target vehicle and the driving orientation of the lane change with the driving moment, where each third coordinate system is a frenet coordinate system, a reference line of each third coordinate system is determined based on a center line of each lane, and an origin of each third coordinate system is determined based on an end point of the center line of each lane”); and
performing a driving maneuver based on the predicted drive intention (Li teaches at least in paragraph [0122] “The prediction unit 43 predicts the behavior intention and the future track of the target vehicle based on current map information and the target information sensed by the sensing unit. The planning unit 44 plans a driving route of the vehicle based on a prediction result of the prediction unit and/or output information of the navigation unit 47. The control unit 45 controls, based on the driving route planned by the planning unit, the vehicle to drive on the planned driving route. The target fusion unit 42, the prediction unit 43, the planning unit 43, and the control unit 45 are all implemented in the processor in FIG. 1 or FIG. 2. In a driving process of the vehicle, an intention of another vehicle is predicted in real time, accurately, and reliably, so that the vehicle can predict a traffic condition in front of the vehicle, and establish a traffic situation around the vehicle”).
Li teaches to obtain a predicted drive intention of the target vehicle at a roadway intersection however Li does not explicitly teach (bolded and italic recitations above) as to the predicted drive intention comprises a turn classification, a flow classification, or both (which are maneuver model of the vehicles). However, it is known in the art before the effective filing date of the claimed invention that to determine the predicted drive intention comprises a turn classification, a flow classification, or both. For example, Neil teaches to predict a maneuver that an object in a driving environment of the autonomous vehicle and to determine the predicted drive intention comprises a turn classification (label), a flow classification (label), or both. Neil further implicitly teaches that having such classifications (labels) provides efficiency of generates predictions of maneuvers that are to be undertaken by objects in environments of autonomous vehicles (see at least Neil Figs. 1-2 and 8-9 and paragraphs [0008]-[0009] and [0034]-[0035]). Therefore it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify the system of Li to include the turn classification, a flow classification, or both in predicted drive intention as taught by Neil in order to provide efficiency of generates predictions of maneuvers that are to be undertaken by objects in environments of autonomous vehicles.
In Reference to Claim 22
The method of claim 21 (see rejection to claim 21 above), wherein the one or more agent tensors representing: a plurality of polylines representing trajectories of the target vehicle and one or more neighbor vehicles of the target vehicle over a most recent period of time, a plurality of points of each of the plurality of polylines, and a plurality of features of each of the plurality of polylines (Li teaches at least in paragraph [0149] “a method for obtaining the interaction feature between the target vehicle and the another vehicle may be extracted according to the following rule: extracting one or more of a location feature of each of the surrounding vehicles in a first coordinate system, a speed feature, and a head orientation feature, where an origin of the first coordinate system is a current location of the target vehicle, the first coordinate system is a rectangular coordinate system, a y-axis of the first coordinate system is parallel to a length direction of a vehicle body of the target vehicle, and a forward direction of the y-axis is consistent with a head orientation of the target vehicle”).
In Reference to Claim 23
The method of claim 21 (see rejection to claim 21 above), wherein the one or more map tensors are three-dimensional tensors representing: a plurality of polylines representing lanes, lane boundaries, or both along which the target vehicle and one or more neighbor vehicles are traveling, a plurality of points of each of the plurality of polylines, and a plurality of features of each of the plurality of polylines (Li teaches at least in paragraph [0151] “the interaction feature between the target vehicle and each lane may be extracted according to the following rule: extracting one or more of a location feature of a target vehicle in each third coordinate system, a feature of an angle formed by the head orientation of the target vehicle and a driving direction of the lane, and a feature that a location of the target vehicle in each third coordinate system, and the angle formed by the head orientation of the target vehicle and the driving orientation of the lane change with the driving moment, where each third coordinate system is a frenet coordinate system, a reference line of each third coordinate system is determined based on a center line of each lane, and an origin of each third coordinate system is determined based on an end point of the center line of each lane”).
In Reference to Claim 25
The method of claim 21 (see rejection to claim 21 above), wherein the turn classification represents a probability that the target vehicle will perform one of a plurality of classes of turns (Neil teaches at least in Figs.2 and 9 and paragraph [0008], [0034] and [0062] “the maneuver label is indicative of a maneuver that the object executes during the time period. For instance, the maneuver label may be straight, left lane change, right lane change, left turn, right turn, stationary, or unknown” “As shown in FIG. 2, the maneuver labels 202-214 include a straight label 202, a left lane change label 204, a right lane change label 206, a left turn label 208, a right turn label 210, a stationary label 212, and an unknown label 214. The unknown label 214 captures maneuvers executed by objects that are not captured by the maneuver labels 202-212. In an example, the maneuver label generation application 106 can assign one of the maneuver labels 202-214 to the sensor data 1 216 and one of the maneuver labels 202-214 to the sensor data M 218. Thus, the maneuver label generation application 106 generates labeled sensor data” and “As such, it may be ambiguous as to whether the vehicle 902 is to execute a left hand turn 908, continue a straight heading 910, or execute a (wide) right turn 912. Using the above-described processes, the autonomous vehicle 400 may provide sensor data generated by the sensor systems 402-404 as input to the machine learning model 502. The machine learning model 502 may then output an indication of a maneuver that the vehicle 902 is predicted to execute. For instance, the machine learning model 502 may be configured to output a probability distribution over the left hand turn 908, the straight heading 910, and the right turn 912. The autonomous vehicle 400 may then operate based upon the indication of the maneuver that the vehicle 902 is predicted to execute”).
In Reference to Claim 26
The method of claim 21 (see rejection to claim 21 above), wherein the flow classification represents a probability that the target vehicle will perform one of a plurality of classes of flows (Neil teaches at least in Figs.2 and 9 and paragraph [0008], [0034] and [0062] “the maneuver label is indicative of a maneuver that the object executes during the time period. For instance, the maneuver label may be straight, left lane change, right lane change, left turn, right turn, stationary, or unknown” “As shown in FIG. 2, the maneuver labels 202-214 include a straight label 202, a left lane change label 204, a right lane change label 206, a left turn label 208, a right turn label 210, a stationary label 212, and an unknown label 214. The unknown label 214 captures maneuvers executed by objects that are not captured by the maneuver labels 202-212. In an example, the maneuver label generation application 106 can assign one of the maneuver labels 202-214 to the sensor data 1 216 and one of the maneuver labels 202-214 to the sensor data M 218. Thus, the maneuver label generation application 106 generates labeled sensor data” and “As such, it may be ambiguous as to whether the vehicle 902 is to execute a left hand turn 908, continue a straight heading 910, or execute a (wide) right turn 912. Using the above-described processes, the autonomous vehicle 400 may provide sensor data generated by the sensor systems 402-404 as input to the machine learning model 502. The machine learning model 502 may then output an indication of a maneuver that the vehicle 902 is predicted to execute. For instance, the machine learning model 502 may be configured to output a probability distribution over the left hand turn 908, the straight heading 910, and the right turn 912. The autonomous vehicle 400 may then operate based upon the indication of the maneuver that the vehicle 902 is predicted to execute”).
In Reference to Claim 27
The method of claim 21 (see rejection to claim 21 above), wherein the predicted drive intention comprises a turn trajectory associated with the turn classification (Neil teaches at least in Figs.2 and 9 and paragraph [0008], [0034] and [0062] “the maneuver label is indicative of a maneuver that the object executes during the time period. For instance, the maneuver label may be straight, left lane change, right lane change, left turn, right turn, stationary, or unknown” “As shown in FIG. 2, the maneuver labels 202-214 include a straight label 202, a left lane change label 204, a right lane change label 206, a left turn label 208, a right turn label 210, a stationary label 212, and an unknown label 214. The unknown label 214 captures maneuvers executed by objects that are not captured by the maneuver labels 202-212. In an example, the maneuver label generation application 106 can assign one of the maneuver labels 202-214 to the sensor data 1 216 and one of the maneuver labels 202-214 to the sensor data M 218. Thus, the maneuver label generation application 106 generates labeled sensor data” and “As such, it may be ambiguous as to whether the vehicle 902 is to execute a left hand turn 908, continue a straight heading 910, or execute a (wide) right turn 912. Using the above-described processes, the autonomous vehicle 400 may provide sensor data generated by the sensor systems 402-404 as input to the machine learning model 502. The machine learning model 502 may then output an indication of a maneuver that the vehicle 902 is predicted to execute. For instance, the machine learning model 502 may be configured to output a probability distribution over the left hand turn 908, the straight heading 910, and the right turn 912. The autonomous vehicle 400 may then operate based upon the indication of the maneuver that the vehicle 902 is predicted to execute”).
In Reference to Claim 28
The method of claim 21 (see rejection to claim 21 above), wherein the driving maneuver comprises: a lane change before or after the roadway intersection, a left turn at the roadway intersection, a right turn at the roadway intersection, a U-turn at the roadway intersection, driving straight through the roadway intersection, a merge into a lane on which the target vehicle is driving, or a hard braking event (Neil teaches at least in Figs.2 and 9 and paragraph [0008], [0034] and [0062] “the maneuver label is indicative of a maneuver that the object executes during the time period. For instance, the maneuver label may be straight, left lane change, right lane change, left turn, right turn, stationary, or unknown” “As shown in FIG. 2, the maneuver labels 202-214 include a straight label 202, a left lane change label 204, a right lane change label 206, a left turn label 208, a right turn label 210, a stationary label 212, and an unknown label 214. The unknown label 214 captures maneuvers executed by objects that are not captured by the maneuver labels 202-212. In an example, the maneuver label generation application 106 can assign one of the maneuver labels 202-214 to the sensor data 1 216 and one of the maneuver labels 202-214 to the sensor data M 218. Thus, the maneuver label generation application 106 generates labeled sensor data” and “As such, it may be ambiguous as to whether the vehicle 902 is to execute a left hand turn 908, continue a straight heading 910, or execute a (wide) right turn 912. Using the above-described processes, the autonomous vehicle 400 may provide sensor data generated by the sensor systems 402-404 as input to the machine learning model 502. The machine learning model 502 may then output an indication of a maneuver that the vehicle 902 is predicted to execute. For instance, the machine learning model 502 may be configured to output a probability distribution over the left hand turn 908, the straight heading 910, and the right turn 912. The autonomous vehicle 400 may then operate based upon the indication of the maneuver that the vehicle 902 is predicted to execute”).
In Reference to Claim 29
Li teaches (except for the bolded and italic recitations below):
An ego vehicle, comprising:
means for applying a machine learning model (30) (see at least Li Fig. 3 and paragraph [0106]) to one or more agent tensors (Li teaches at least in paragraph [0160] “the interaction feature between the target vehicle and each of the other vehicles is input into an interaction feature vector extraction network corresponding to the target vehicle and the other vehicle, to extract an interaction feature vector between the target vehicle and the other vehicle. The vector represents impact of the other vehicle on the driving intention of the target vehicle”) and one or more map tensors (Li teaches at least in paragraph [0171] “The interaction feature between the target vehicle and the road associated with each lane is input into the first road feature extraction subnetwork, to extract an interaction feature vector between the target vehicle and the road associated with each lane; and the interaction feature vector between the target vehicle and the road associated with each lane is input into the second road feature extraction subnetwork, to extract an interaction feature implicit vector between the target vehicle and the road associated with each lane”) associated with a target vehicle to obtain a predicted drive intention of the target vehicle at a roadway intersection (see Li Figs. 6-8), wherein the predicted drive intention comprises a turn classification, a flow classification, or both (see at least Li Fig. 3 and paragraph [0106]), and wherein the one or more agent tensors are three-dimensional tensors (Li teaches at least in paragraphs [0149], [0150] and [0151] “a method for obtaining the interaction feature between the target vehicle and the another vehicle may be extracted according to the following rule: extracting one or more of a location feature of each of the surrounding vehicles in a first coordinate system, a speed feature, and a head orientation feature, where an origin of the first coordinate system is a current location of the target vehicle, the first coordinate system is a rectangular coordinate system, a y-axis of the first coordinate system is parallel to a length direction of a vehicle body of the target vehicle, and a forward direction of the y-axis is consistent with a head orientation of the target vehicle”, “In step S503, a driving feature of the target vehicle relative to each lane, namely, an interaction feature between the target vehicle and each lane, is determined based on the driving information of the target vehicle and the lane layer information” and “the interaction feature between the target vehicle and each lane may be extracted according to the following rule: extracting one or more of a location feature of a target vehicle in each third coordinate system, a feature of an angle formed by the head orientation of the target vehicle and a driving direction of the lane, and a feature that a location of the target vehicle in each third coordinate system, and the angle formed by the head orientation of the target vehicle and the driving orientation of the lane change with the driving moment, where each third coordinate system is a frenet coordinate system, a reference line of each third coordinate system is determined based on a center line of each lane, and an origin of each third coordinate system is determined based on an end point of the center line of each lane”); and
means for performing a driving maneuver based on the predicted drive intention (Li teaches at least in paragraph [0122] “The prediction unit 43 predicts the behavior intention and the future track of the target vehicle based on current map information and the target information sensed by the sensing unit. The planning unit 44 plans a driving route of the vehicle based on a prediction result of the prediction unit and/or output information of the navigation unit 47. The control unit 45 controls, based on the driving route planned by the planning unit, the vehicle to drive on the planned driving route. The target fusion unit 42, the prediction unit 43, the planning unit 43, and the control unit 45 are all implemented in the processor in FIG. 1 or FIG. 2. In a driving process of the vehicle, an intention of another vehicle is predicted in real time, accurately, and reliably, so that the vehicle can predict a traffic condition in front of the vehicle, and establish a traffic situation around the vehicle”).
Li teaches to obtain a predicted drive intention of the target vehicle at a roadway intersection however Li does not explicitly teach (bolded and italic recitations above) as to the predicted drive intention comprises a turn classification, a flow classification, or both (which are maneuver model of the vehicles). However, it is known in the art before the effective filing date of the claimed invention that to determine the predicted drive intention comprises a turn classification, a flow classification, or both. For example, Neil teaches to predict a maneuver that an object in a driving environment of the autonomous vehicle and to determine the predicted drive intention comprises a turn classification (label), a flow classification (label), or both. Neil further implicitly teaches that having such classifications (labels) provides efficiency of generates predictions of maneuvers that are to be undertaken by objects in environments of autonomous vehicles (see at least Neil Figs. 1-2 and 8-9 and paragraphs [0008]-[0009] and [0034]-[0035]). Therefore it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify the system of Li to include the turn classification, a flow classification, or both in predicted drive intention as taught by Neil in order to provide efficiency of generates predictions of maneuvers that are to be undertaken by objects in environments of autonomous vehicles.
In Reference to Claim 30
Li teaches (except for the bolded and italic recitations below):
A non-transitory computer-readable medium storing computer-executable instructions that, when executed by an ego vehicle, cause the ego vehicle to:
apply a machine learning model (30) (see at least Li Fig. 3 and paragraph [0106]) to one or more agent tensors (Li teaches at least in paragraph [0160] “the interaction feature between the target vehicle and each of the other vehicles is input into an interaction feature vector extraction network corresponding to the target vehicle and the other vehicle, to extract an interaction feature vector between the target vehicle and the other vehicle. The vector represents impact of the other vehicle on the driving intention of the target vehicle”) and one or more map tensors (Li teaches at least in paragraph [0171] “The interaction feature between the target vehicle and the road associated with each lane is input into the first road feature extraction subnetwork, to extract an interaction feature vector between the target vehicle and the road associated with each lane; and the interaction feature vector between the target vehicle and the road associated with each lane is input into the second road feature extraction subnetwork, to extract an interaction feature implicit vector between the target vehicle and the road associated with each lane”) associated with a target vehicle to obtain a predicted drive intention of the target vehicle at a roadway intersection (see Li Figs. 6-8), wherein the predicted drive intention comprises a turn classification, a flow classification, or both (see at least Li Fig. 3 and paragraph [0106]), and wherein the one or more agent tensors are three-dimensional tensors (Li teaches at least in paragraphs [0149], [0150] and [0151] “a method for obtaining the interaction feature between the target vehicle and the another vehicle may be extracted according to the following rule: extracting one or more of a location feature of each of the surrounding vehicles in a first coordinate system, a speed feature, and a head orientation feature, where an origin of the first coordinate system is a current location of the target vehicle, the first coordinate system is a rectangular coordinate system, a y-axis of the first coordinate system is parallel to a length direction of a vehicle body of the target vehicle, and a forward direction of the y-axis is consistent with a head orientation of the target vehicle”, “In step S503, a driving feature of the target vehicle relative to each lane, namely, an interaction feature between the target vehicle and each lane, is determined based on the driving information of the target vehicle and the lane layer information” and “the interaction feature between the target vehicle and each lane may be extracted according to the following rule: extracting one or more of a location feature of a target vehicle in each third coordinate system, a feature of an angle formed by the head orientation of the target vehicle and a driving direction of the lane, and a feature that a location of the target vehicle in each third coordinate system, and the angle formed by the head orientation of the target vehicle and the driving orientation of the lane change with the driving moment, where each third coordinate system is a frenet coordinate system, a reference line of each third coordinate system is determined based on a center line of each lane, and an origin of each third coordinate system is determined based on an end point of the center line of each lane”); and
perform a driving maneuver based on the predicted drive intention (Li teaches at least in paragraph [0122] “The prediction unit 43 predicts the behavior intention and the future track of the target vehicle based on current map information and the target information sensed by the sensing unit. The planning unit 44 plans a driving route of the vehicle based on a prediction result of the prediction unit and/or output information of the navigation unit 47. The control unit 45 controls, based on the driving route planned by the planning unit, the vehicle to drive on the planned driving route. The target fusion unit 42, the prediction unit 43, the planning unit 43, and the control unit 45 are all implemented in the processor in FIG. 1 or FIG. 2. In a driving process of the vehicle, an intention of another vehicle is predicted in real time, accurately, and reliably, so that the vehicle can predict a traffic condition in front of the vehicle, and establish a traffic situation around the vehicle”).
Li teaches to obtain a predicted drive intention of the target vehicle at a roadway intersection however Li does not explicitly teach (bolded and italic recitations above) as to the predicted drive intention comprises a turn classification, a flow classification, or both (which are maneuver model of the vehicles). However, it is known in the art before the effective filing date of the claimed invention that to determine the predicted drive intention comprises a turn classification, a flow classification, or both. For example, Neil teaches to predict a maneuver that an object in a driving environment of the autonomous vehicle and to determine the predicted drive intention comprises a turn classification (label), a flow classification (label), or both. Neil further implicitly teaches that having such classifications (labels) provides efficiency of generates predictions of maneuvers that are to be undertaken by objects in environments of autonomous vehicles (see at least Neil Figs. 1-2 and 8-9 and paragraphs [0008]-[0009] and [0034]-[0035]). Therefore it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify the system of Li to include the turn classification, a flow classification, or both in predicted drive intention as taught by Neil in order to provide efficiency of generates predictions of maneuvers that are to be undertaken by objects in environments of autonomous vehicles.
Claims 4 and 7 are rejected under 35 U.S.C. 103 as being unpatentable over Li in view of Neil and further in view of Pub No. CN 113511204 A to Zhang et. al. (Zhang).
Examiner’s Note: Machine translation of Pub No. CN 113511204 A will be used in the rejection below.
In Reference to Claim 4
Li in view of Neil teaches (except for the bolded and italic recitations below):
The ego vehicle of claim 3 (see rejection to claim 3 above), wherein the length of the most recent period of time is one second (see at least Li Fig. 3 and paragraph [0106]).
Li in view of Neil do not teach (bolded and italic recitations above) as to the most recent period of time is one second. However, it is known in the art before the effective filing date of the claimed invention to have the length of the most recent period of time is one second. For example, Zhang teaches the length of the most recent period of time is one second. Zhang further teaches that performing such step provides quickly and accurately determine the possible presence domain of the target vehicle lane changing behavior, so as to provide preliminary information for further detection of subsequent lane changing behavior, and improve the correct rate of vehicle lane changing behavior identification (see at least Zhang page 5). Therefore it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify the system of Li in view of Neil to include the length of the most recent period of time is one second as taught by Zhang in order to provide quickly and accurately determine the possible presence domain of the target vehicle lane changing behavior, so as to provide preliminary information for further detection of subsequent lane changing behavior, and improve the correct rate of vehicle lane changing behavior identification.
In Reference to Claim 7
Li in view of Neil teaches (except for the bolded and italic recitations below):
The ego vehicle of claim 5 (see rejection to claim 5 above), wherein a number of the plurality of points is 20 points (see at least Li Fig. 3 and paragraph [0106]).
Li in view of Neil do not teach (bolded and italic recitations above) as to a number of the plurality of points is 20 points. However, it is known in the art before the effective filing date of the claimed invention to have a number of the plurality of points is 20 points. For example, Zhang teaches a number of the plurality of points is 20 points. Zhang further teaches that performing such step provides quickly and accurately determine the possible presence domain of the target vehicle lane changing behavior, so as to provide preliminary information for further detection of subsequent lane changing behavior, and improve the correct rate of vehicle lane changing behavior identification (see at least Zhang page 5). Therefore it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify the system of Li in view of Neil to include the length of the most recent period of time is one second as taught by Zhang in order to provide quickly and accurately determine the possible presence domain of the target vehicle lane changing behavior, so as to provide preliminary information for further detection of subsequent lane changing behavior, and improve the correct rate of vehicle lane changing behavior identification.
Claims 9-14 and 24 are rejected under 35 U.S.C. 103 as being unpatentable over Li in view of Neil and further in view of Pub No. US 2024/0281988 A1 to Abbott et. al. (Abbott).
In Reference to Claim 9
Li in view of Neil teaches (except for the bolded and italic recitations below):
The ego vehicle of claim 1 (see rejection to claim 1 above), wherein the machine learning model is an encoder-decoder machine learning model (see at least Li Figs. 1-3 and paragraph [0017]-[0018] and [0034]).
Li in view of Neil do not explicitly teaches (bolded and italic recitations above) that the machine learning model is an encoder-decoder machine learning model. However, it is known in the art before the effective filing date of the claimed invention to that the machine learning model is an encoder-decoder machine learning model. For example, Abbott teaches the machine learning model is an encoder-decoder machine learning model (see at least Abbott Figs. 1-2 and paragraphs 7-8, 29-31, 46-59, 62 and 64).The substitution of one known element (an encoder-decoder machine learning model as shown in Abbott) for another (machine learning model as shown in Li in view of Neil) would have been obvious to one of ordinary skill in the before the filing date of the claimed invention since the substitution of the machine learning model shown in Abbott would have yielded predictable results, namely, controlling the autonomous driving of the system of Li in view of Neil.
In Reference to Claim 10
The ego vehicle of claim 9 (see rejection to claim 9 above), wherein an encoder side of the encoder-decoder machine learning model comprises a first stage and a second stage (Abbott teaches in paragraph [0008] “multiple DNNs (e.g., a chain of multiple DNNs or multiple stages of a DNN) are used to sequentially generate classifications of measured 3D points and a regressed representation of the shape of one or more detected landmarks. For example, a first (stage of a) DNN may be used to extract classification data (e.g., confidence maps for any number of classes) from any suitable input (e.g., a projection image, a multi-channel projection image or tensor, etc.), and a second (stage of the) DNN may be used to extract shape regression data representing one or more fitted 2D or 3D shapes represented in the classification data. In some such embodiments, the classification DNN (or stage) operates similarly as the classification head described above, and the regression DNN (or stage) operates similarly as the regression head described above, except the two DNNs (or stages) are serialized rather than output from the same encoder/decoder trunk of the DNN”) (see at least Abbott Figs. 1-2 and paragraphs 7-8, 29-31, 46-59, 62 and 64).
In Reference to Claim 11
The ego vehicle of claim 10 (see rejection to claim 10 above), wherein the first stage comprises: an agent polyline module applied to the one or more agent tensors, wherein the agent polyline module converts the one or more agent tensors to one or more two-dimensional agent tensors, and a map polyline module applied to the one or more map tensors, wherein the map polyline module converts the one or more map tensors to one or more two-dimensional map tensors (Abbott teaches in paragraph [0029] and [0062] “a representation of the shape of one or more detected landmarks is regressed from the classifications (e.g., performing a connected components on a two-dimensional (2D) representation of the classifications and fitting one or more shapes to connected components in the 2D representation, or using a DNN to regress a representation of one or more fitted polylines or circles). In some embodiments, a DNN is used to jointly generate classifications of measured 3D points from one output head (e.g., a classification head) and regress a representation of one or more fitted shapes (e.g., polylines, circles) from a second output head (e.g., a regression head)” and “For example and by way of illustration, if the goal is to regress 2D/3D polylines fitted to classes of landmarks such as lane lines, road boundaries, and/or poles in perspective view, the 2D/3D polylines may be parameterized into Bezier splines and/or n-degree polynomials. FIG. 8 is an illustration of example shape regression data that may be predicted by one or more machine learning models, in accordance with some embodiments of the present disclosure. FIG. 8 illustrates a grid where each cell in the grid (e.g., cell 805) is illustrated as a dot and represents a candidate anchor point for regressed polylines in perspective view, dotted lines represent ground truth lane lines, the enlarged dots (e.g., point 810) represent predicted anchor points with predicted confidence levels above a threshold, and solid lines represent predicted polylines parametrized by predicted parameters associated with the predicted anchor points. In some embodiments, classification need not be on a per-LiDAR point basis. For example, pixels and/or anchor points (e.g., and therefore, the class confidence data 410 and/or the shape regression data 412) may be predicted at some lower resolution (e.g., 400×200 cells)”).
In Reference to Claim 12
The ego vehicle of claim 11 (see rejection to claim 11 above), wherein the first stage further comprises: a self-attention module applied to the one or more two-dimensional agent tensors and the one or more two-dimensional map tensors to obtain a one-dimensional vector representing the predicted drive intention of the target vehicle (Abbott teaches in paragraph [0068] and [0072] “each control point may be parameterized with two dimensions (e.g., x, y), each of which may be assigned and encoded into to one or more corresponding regression channels. As such, ground truth shape regression data may be generated with the same size and/or and dimensionality as the shape regression data 412 of FIG. 4” and “The multi-class cross-entropy loss may include a pixel-wise cross-entropy loss summed over all classes. For example, classification loss may be given by H(p, y)=−Σ.sub.iy.sub.i log(p.sub.i), where y is ground truth data (e.g., 1 where a pixel is in the class, 0 otherwise), p is a predicted depth-wise pixel vector with each dimension corresponding to a particular class, p is the predicted classification data per pixel (e.g., probability, score, or logit that the pixel is in a given class i), and the summation may be performed per pixel over all classes (e.g., over all depth channels). Classification loss for a full frame may be computed by computing classification loss for each pixel and taking its mean value”).
In Reference to Claim 13
The ego vehicle of claim 10 (see rejection to claim 10 above), wherein the second stage comprises: a shared fully connected multi-layer perception (MLP) layer, and one or more individual fully connected MLP layers for the turn classification and the flow classification (Abbott teaches in paragraph [0089] and [0091] “As such, the obstacle avoidance layer may be a separate layer from the rules of the road layer, and the obstacle avoidance layer may ensure that the vehicle 1200 is only performing safe actions from an obstacle avoidance standpoint. The rules of the road layer, on the other hand, may ensure that vehicle obeys traffic laws and conventions, and observes lawful and conventional right of way (as described herein)” and “In non-limiting embodiments, the obstacle avoidance component(s) 132 may be implemented as a separate, discrete feature of the vehicle 1200. For example, the obstacle avoidance component(s) 132 may operate separately (e.g., in parallel with, prior to, and/or after) the planning layer, the control layer, the actuation layer, and/or other layers of the drive stack 122”).
In Reference to Claim 14
The ego vehicle of claim 10 (see rejection to claim 10 above), wherein: the predicted drive intention further comprises a turn trajectory, and the second stage comprises: a shared fully connected multi-layer perception (MLP) layer, and one or more individual fully connected MLP layers for each of a plurality of turn trajectories (Abbott teaches in paragraph [0089] and [0091] “As such, the obstacle avoidance layer may be a separate layer from the rules of the road layer, and the obstacle avoidance layer may ensure that the vehicle 1200 is only performing safe actions from an obstacle avoidance standpoint. The rules of the road layer, on the other hand, may ensure that vehicle obeys traffic laws and conventions, and observes lawful and conventional right of way (as described herein)” and “In non-limiting embodiments, the obstacle avoidance component(s) 132 may be implemented as a separate, discrete feature of the vehicle 1200. For example, the obstacle avoidance component(s) 132 may operate separately (e.g., in parallel with, prior to, and/or after) the planning layer, the control layer, the actuation layer, and/or other layers of the drive stack 122”).
In Reference to Claim 24
Li in view of Neil teaches (except for the bolded and italic recitations below):
The method of claim 21 (see rejection to claim 21 above), wherein the machine learning model is an encoder-decoder machine learning model (see at least Li Figs. 1-3 and paragraph [0017]-[0018] and [0034]).
Li in view of Neil do not explicitly teaches (bolded and italic recitations above) that the machine learning model is an encoder-decoder machine learning model. However, it is known in the art before the effective filing date of the claimed invention to that the machine learning model is an encoder-decoder machine learning model. For example, Abbott teaches the machine learning model is an encoder-decoder machine learning model (see at least Abbott Figs. 1-2 and paragraphs 7-8, 29-31, 46-59, 62 and 64).The substitution of one known element (an encoder-decoder machine learning model as shown in Abbott) for another (machine learning model as shown in Li in view of Neil) would have been obvious to one of ordinary skill in the before the filing date of the claimed invention since the substitution of the machine learning model shown in Abbott would have yielded predictable results, namely, controlling the autonomous driving of the system of Li in view of Neil.
Claim 16 rejected under 35 U.S.C. 103 as being unpatentable over Li in view of Neil and further in view of Pub No. US 2022/0410938 A1 to Huang et. al. (Huang).
In Reference to Claim 16
Li in view of Neil teaches (except for the bolded and italic recitations below):
The ego vehicle of claim 15 (see rejection to claim 15 above), wherein the plurality of classes of turns comprises: left turn, right turn, straight, and U-turn (see at least Neil Figs. 1-2 and 8-9 and paragraphs [0008]-[0009] and [0034]-[0035]).
Li in view of Neil do not explicitly teaches (bolded and italic recitations above) as to the plurality of classes of turns comprises a U-turn. However, it is known in the art before the effective filing date of the claimed invention to plurality of classes of turns comprises a U-turn. For example, Huang teaches plurality of classes of turns comprises a U-turn. Further Huang implicitly teaches that U-turn is one of the modes that vehicle can operate in the intersection (see at least Huang paragraph [0027]). Therefore it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify the system of Li in view of Neil to include the U-turn in the plurality of classes of turns as taught by Huang in order to include all the modes that vehicle can operate in the intersection.
Response to Arguments
Applicant's arguments filed 3/18/2026 have been fully considered but they are not persuasive.
The applicant argues that Li do not teach “In the recited portions of Li above, Li does not mention or suggest any three-dimensional agent tensor. Various features that are mentioned by Li (e.g., interaction feature, location feature, speed feature, head orientation feature, etc.) are not three-dimensional agent tensor as claimed in claim 1. In fact, Li does not mention any three-dimensional structure representing any feature that is disclosed (e.g., driving feature, interaction feature, location feature, etc.). Li only mentions feature vectors (e.g., driving feature vector, interaction feature vector) but does not mention or suggest that any of these feature vectors is a three-dimensional tensor as claimed in claim 1. The various features or feature vectors mentioned in Li are not the same and are different from the three-dimensional agent tensor claimed in claim 1. Other references cited above do not cure the deficiencies of Li because these references, either alone or in combinations, do not disclose or suggest the above-recited limitations of claim 1 either. Accordingly, Applicants respectfully submit that claim 1, as amended, is patentable over Li and the references cited above, either alone or in combinations. Independent claim 21, 29 and 30 include similar features and limitations and therefore are also patentable over Li and the references cited above. Their respective dependent claims are therefore also patentable at least by the virtue of their dependencies” however the examiner respectfully disagree with the applicant since Li do teaches three-dimensional tensor since interaction feature, location feature, speed feature, head orientation feature, etc. can be interpreted as three-dimensional tensor since the claim 1 do not explicitly discloses as to what comprises the three-dimensional tensor are and further Li teaches (e.g., driving feature, interaction feature, location feature, etc.) in paragraphs 150-151 as shown in the rejections to claims 1, 21 and 29-30.
Conclusion
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure.
Pub No. US 2021/0001844 A1 to Perincherry et. al. (Perincherry) teaches vehicle intersection operation of autonomous vehicle.
THIS ACTION IS MADE FINAL. Applicant is reminded of the extension of time policy as set forth in 37 CFR 1.136(a).
A shortened statutory period for reply to this final action is set to expire THREE MONTHS from the mailing date of this action. In the event a first reply is filed within TWO MONTHS of the mailing date of this final action and the advisory action is not mailed until after the end of the THREE-MONTH shortened statutory period, then the shortened statutory period will expire on the date the advisory action is mailed, and any nonprovisional extension fee (37 CFR 1.17(a)) pursuant to 37 CFR 1.136(a) will be calculated from the mailing date of the advisory action. In no event, however, will the statutory period for reply expire later than SIX MONTHS from the mailing date of this final action.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to BRANDON DONGPA LEE whose telephone number is (571)270-3525. The examiner can normally be reached Monday - Friday, 8:00 am - 5:00 pm.
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, Aniss Chad can be reached at (571) 270-3832. 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.
/BRANDON D LEE/Primary Examiner, Art Unit 3662 May 1, 2026
Prosecution Insights