DETAILED ACTION
Notice of Pre-AIA or AIA Status
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
Continued Examination Under 37 CFR 1.114
A request for continued examination under 37 CFR 1.114, including the fee set forth in 37 CFR 1.17(e), was filed in this application after final rejection. Since this application is eligible for continued examination under 37 CFR 1.114, and the fee set forth in 37 CFR 1.17(e) has been timely paid, the finality of the previous Office action has been withdrawn pursuant to 37 CFR 1.114. Applicant's submission filed on 06/29/2026 has been entered.
Priority
Receipt is acknowledged of certified copies of papers required by 37 CFR 1.55. Priority is being given to 03/15/2023.
Status of Claims
This action is in reply to the amendments filed on 06/29/2026.
Claims 1-20 are currently pending and have been examined.
Claims 1, 3-4, 13, and 15-16 are amended.
Claims 1-20 are currently rejected.
This action is made NON-FINAL.
Response to Arguments
Applicant’s arguments filed 03/29/2026 have been fully considered but are not persuasive.
Applicant’s arguments with regards to the art rejections have been considered and are not persuasive. Applicant argues that art of record does not teach the amended limitations. The rejections have been updated in light of the amendments.
Specification
The disclosure is objected to because of the following informalities: the specification in various spots recites “encoder module”, “encoder-decoder model”, and “decoder model”. It is appears “model” is a typo for “module”. Appropriate correction is required.
Claim Rejections - 35 USC § 112
The following is a quotation of 35 U.S.C. 112(b):
(b) CONCLUSION.—The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the inventor or a joint inventor regards as the invention.
The following is a quotation of 35 U.S.C. 112 (pre-AIA ), second paragraph:
The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the applicant regards as his invention.
Claims 1 and 13 recites the limitations “encoder module”, “encoder-decoder model”, and “decoder model”. It is unclear if the inventor is intending to claim a module or a model for each of these elements. To overcome this rejection the examiner suggest changing “encoder-decoder model” and “decoder model” to “encoder-decoder module” and “decoder module”. For the purposes of examination, the examiner is interpreting the recitations to be referring to a “module”.
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.
The factual inquiries for establishing a background for determining obviousness under 35 U.S.C. 103 are summarized as follows:
1. Determining the scope and contents of the prior art.
2. Ascertaining the differences between the prior art and the claims at issue.
3. Resolving the level of ordinary skill in the pertinent art.
4. Considering objective evidence present in the application indicating obviousness or nonobviousness.
Claim(s) 1-4, 9-16, and 20 is/are rejected under 35 U.S.C. 102(a)(2) as being anticipated by Li et. al. (US 2023/0399023), herein Li in view of Sun et. al. (US 2021/0174668), herein Sun.
Regarding claim 1:
Li teaches:
A method for predicting a vehicle trajectory (a vehicle driving intention prediction method [0006]), comprising:
obtaining vectorized features of a plurality of target vehicles (to determine an interaction feature vector between the surrounding vehicle and the target vehicle, where the interaction feature vector between the surrounding vehicle and the target vehicle represents impact of the surrounding vehicle on the target vehicle [0010]; a driving feature vector of the target vehicle relative to each of the plurality of lanes [0013]) based on perceived information (The sensor system 104 may include several sensors that can sense information about the ambient environment of the vehicle 100. [0062]) of an autonomous vehicle (vehicle 100 operates in an autonomous mode [0084]), wherein the perceived information is obtained based on data acquired by a sensor of the autonomous vehicle (the sensor system 104 may include a positioning system 122 (the positioning system may be a Global Positioning System (GPS), a BeiDou system, or another positioning system), an inertial measurement unit (IMU) 124, a radar 126, a laser rangefinder 128, and a camera 130. The sensor system 104 may further include sensors (for example, an in-vehicle air quality monitor, a fuel gauge, and an oil temperature gauge) in an internal system of the vehicle 100. Sensor data from one or more of these sensors can be used to detect an object and corresponding features (a location, a shape, a direction, a speed, and the like). Such detection and recognition are key functions of safe operation of the vehicle 100 [0063]);
obtaining a trajectory prediction result (The prediction unit 43 predicts the behavior intention and the future track of the target vehicle [0122]) based on the encoded features of each target vehicle of the plurality of target vehicles (based on current map information and the target information sensed by the sensing unit [0122]), to obtain a plurality of trajectory prediction results of the plurality of target vehicles (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. This helps determine importance of the target of another surrounding vehicle of the vehicle, and filter a key target for interaction, so that the vehicle can plan the route in advance and safely pass through a complex road condition scenario. [0122]), wherein the plurality of target vehicles (extracting one or more of a location feature of each of the surrounding vehicles [0149]) comprise the autonomous vehicle (fig. 6, current vehicle) and a plurality of first surrounding vehicles (fig. 6, another vehicle), wherein the plurality of first surrounding vehicles being vehicles in a surrounding environment of the autonomous vehicle that are selected according to a preset first rule (The surrounding vehicle of the target vehicle may be understood as another vehicle that is at a specific distance from the target vehicle. The distance may be set by a user, or may be set by a skilled person, or may be related to a sensing distance of a sensor of the current vehicle. [0145]);
Controlling the autonomous vehicle based on the trajectory prediction result obtained (the application 141 may also be a program for controlling the autonomous vehicle to avoid collision with another vehicle and safely pass through an intersection [0102]).
Sun also teaches:
A method for predicting a vehicle trajectory (it is desirable to consider the traffic lanes that surround a target vehicle in order to predict the vehicle's future movements [0024]), comprising:
obtaining vectorized features of a plurality of target vehicles based on perceived information (the one or more sensors 102 installed in the target vehicle 100 captures velocity and position information of the nearby vehicles 200, as well as the velocity and position information of the target vehicle 100. The position and velocity information may be captured over multiple time steps, and used by the perception module 108 for determining a trajectory 202 of the target vehicle 100 and the nearby vehicles 200 [0042]) of an autonomous vehicle (A self-driving vehicle (autonomous vehicle) [0003]), wherein the perceived information is obtained based on data acquired by a sensor of the autonomous vehicle (the one or more sensors 102 of the target vehicle 100 capture and provide sensor data of nearby objects [0041]);
a trajectory prediction result (output a future trajectory predicted for the target vehicle 100 [0051]) based on the encoded vectorized features of each target vehicle of the plurality of target vehicles (The trajectory encoder 208 may include one or more be neural networks (e.g. LSTM, GRU, or the like) trained to extract and encode features of the trajectories for the target vehicle 100 and nearby vehicles 200. In one embodiment, the trajectory encoder 208 includes a first neural network for encoding position histories of an agent into a first vector, and a second neural network for encoding velocity histories of the agent in a second vector. [0045]), to obtain a plurality of trajectory prediction results of the plurality of target vehicles (The interaction network module 214 may be configured to model the interactions between the various agents 100, 200. In some situations, the future trajectory of the target vehicle may not only be influenced by the lanes that are subject to the driver's attention, but may also be influenced by interactions with the other vehicles 200 [0049]), wherein the plurality of target vehicles comprise the autonomous vehicle (fig. 2, target vehicle 100) and a plurality of first surrounding vehicles (fig. 2, other vehicles 200a-200c),
controlling the autonomous vehicle based on the trajectory prediction result obtained (the planning module 111 of the target vehicle may consider the various trajectories in planning a path for the target vehicle [0066]).
Li does not explicitly teach, however Sun teaches:
inputting, into an encoder module (fig. 2, trajectory encoder 208) of an encoder-decoder network model (fig. 2, prediction module 110), the vectorized features of the plurality of target vehicles to obtain encoded features of each target vehicle of the plurality of target vehicles (The trajectory encoder 208 may include one or more be neural networks (e.g. LSTM, GRU, or the like) trained to extract and encode features of the trajectories for the target vehicle 100 and nearby vehicles 200 [0045]);
obtaining, by a decoder model (fig. 2, decoder 216) of the encoder-decoder network model (the decoder module 216 is a standard LSTM network or any other recurrent neural network. The decoder module 216 may be configured to receive the trajectory feature of the target vehicle 100, the interaction features generated by the interaction network module 214, and the feature of the target lane generated by the lane attention module 212 [0051]),
It would have been obvious to one of ordinary skill in the art at the time of the effective filing date of the claimed invention to have modified Li to include the teachings as taught by Sun with a reasonable expectation of success. Both arts are in the same field of endeavor of autonomous vehicle control. Sun teaches the benefits of “A self-driving vehicle (autonomous vehicle) engages in various tasks to safely maneuver itself in a current environment. One of such tasks is predicting its future trajectory as well as the trajectory of surrounding vehicles. Trajectory prediction, however, may often be challenging due to the inherent uncertainty of the future, and the complexity of the environment navigated by the vehicle. Trajectory prediction may also be challenging as the driver's intention often determines trajectory, and intention may be hard to estimate. It is desirable, therefore, to have a system and method that tackles the challenging task of trajectory prediction for autonomous vehicles [Sun, 0003]”.
Regarding claim 2:
Li in view of Sun teaches all the limitations of claim 1, upon which this claim is dependent.
Li further teaches:
wherein obtaining vectorized features of the plurality of target vehicles (a driving feature vector of the target vehicle relative to each of the plurality of lanes [0013]) further comprises:
obtaining, based on the perceived information of the autonomous vehicle (the processor 113 may predict a driving track of another vehicle based on a surrounding road condition and another vehicle condition that are detected by the sensor 153 [0104]), self- vehicle trajectory vectorized features (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 [0122]), surrounding trajectory vectorized features (the sensor 153 may detect an animal, a vehicle, an obstacle, or cross walk [0103]), and road network vectorized features of each target vehicle of the plurality of target vehicles (a driving feature implicit vector of the target vehicle relative to each of the plurality of lanes [0010]);
wherein the surrounding trajectory vectorized features are trajectory vectorized features of a plurality of second surrounding vehicles of a target vehicle of the plurality of target vehicles (a driving feature vector of each of the surrounding vehicles relative to the target vehicle [0012]), the plurality of second surrounding vehicles being vehicles in a surrounding environment of the target vehicle that are selected according to a preset second rule (The surrounding vehicle of the target vehicle may be understood as another vehicle that is at a specific distance from the target vehicle. The distance may be set by a user, or may be set by a skilled person, or may be related to a sensing distance of a sensor of the current vehicle. [0145]).
Regarding claim 3:
Li in view of Sun teaches all the limitations of claim 2, upon which this claim is dependent.
Li further teaches:
wherein obtaining the trajectory prediction result (The prediction unit 43 predicts the behavior intention and the future track of the target vehicle [0122]) based on the encoded features of each target vehicle of the plurality of target vehicles (based on current map information and the target information sensed by the sensing unit [0122]) further comprises:
obtaining, for each target vehicle of the plurality of target vehicles, encoded (The target fusion unit 42 processes the environment information around the vehicle sensed by the sensing unit 41, and outputs obstacle target information [0122]) features of the target vehicle (obtaining driving information of the target vehicle [0007]) based on the self-vehicle trajectory vectorized features (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 [0122]), the surrounding trajectory vectorized features (the lane intention of the target vehicle based on the driving feature of the surrounding vehicle relative to the target vehicle [0007]), and the road network vectorized features of the target vehicle (the driving feature of the target vehicle relative to each of the plurality of roads [0008]).
Sun also further teaches:
wherein obtaining the trajectory prediction result (predicting a future trajectory of a target vehicle [0038]) based on the encoded features of each target vehicle of the plurality of target vehicles (The lane feature vectors for the lanes generated by the lane encoder 210, and the trajectory feature vector for the target vehicle 100 that is generated by the trajectory encoder 208, are used by the lane attention module 212 for identifying one or more target lanes that are predicted to be the goal of the target vehicle in the upcoming future (e.g. in the next 3 seconds) [0047]) further comprises:
obtaining, for each target vehicle of the plurality of target vehicles, encoded () features of the target vehicle (a trajectory 202 of the target vehicle 100 and the nearby vehicles 200 [0042]) based on the self-vehicle trajectory vectorized features (a trajectory 202 of the target vehicle 100 [0042]), the surrounding trajectory vectorized features (extracting lane features based on the set of ordered lane-points 204 of the lane segments. The final output of the convolutional neural network may be pooled and stored in a fixed size feature vector for each lane [0046]), and the road network vectorized features of the target vehicle (the perception module 108 is further configured to obtain a local map including a set of ordered lane-points (e.g. geographic x, y coordinates) 204 of the center of the lanes within a threshold vicinity/distance 206 of the target vehicle 100 [0043]).
Regarding claim 4:
Li in view of Sun teaches all the limitations of claim 3, upon which this claim is dependent.
Li further teaches:
wherein the obtaining the encoded features of the target vehicle based on the self-vehicle trajectory vectorized features (The target fusion unit 42 processes the environment information around the vehicle sensed by the sensing unit 41, and outputs obstacle target information [0122]), the surrounding trajectory vectorized features (the lane intention of the target vehicle based on the driving feature of the surrounding vehicle relative to the target vehicle [0007]), and the road network vectorized features of the target vehicle (the driving feature of the target vehicle relative to each of the plurality of roads [0008]) further comprises:
encoding the self-vehicle trajectory vectorized features, the surrounding trajectory vectorized features, and the road network vectorized features (The target fusion unit 42 processes the environment information around the vehicle sensed by the sensing unit 41, and outputs obstacle target information [0122]) to obtain self-vehicle trajectory encoded features, surrounding trajectory encoded features, and road network encoded features (The target fusion unit 42 processes the environment information around the vehicle sensed by the sensing unit 41, and outputs obstacle target information [0122]), respectively;
performing feature interaction (an interaction feature between the target vehicle and another vehicle, may be determined based on the driving information of the target vehicle and the driving information of the surrounding vehicle of the target vehicle [0148]) on the self-vehicle trajectory encoded features, the surrounding trajectory encoded features, and the road network encoded (The target fusion unit 42 processes the environment information around the vehicle sensed by the sensing unit 41, and outputs obstacle target information [0122]) features to obtain self-vehicle trajectory interaction features (the interaction feature between the target vehicle and the other vehicle [0157]), surrounding trajectory interaction features (the interaction feature between the target vehicle and each road [0157]), and environment interaction features (the interaction feature between the target vehicle and each lane [0157]), respectively ; and
performing feature fusion (The target fusion unit 42 processes the environment information around the vehicle sensed by the sensing unit 41 [0122]) on the self-vehicle trajectory interaction features, the surrounding trajectory interaction features, and the environment interaction features to obtain the encoded features of the target vehicle (The driving feature vector of each of the other vehicles relative to the target vehicle is input into the interaction feature vector prediction network, to obtain an interaction feature vector between the another vehicle and the target vehicle. [0161]).
Regarding claim 9:
Li in view of Sun teaches all the limitations of claim 1, upon which this claim is dependent.
Li further teaches:
for each target vehicle of the plurality of target vehicles, structuring the perceived information to obtain structured data of the target vehicle (the driving information of the target vehicle or the surrounding vehicle includes information that can be sensed by the current vehicle and that affects a driving intention of the target vehicle, for example, location information, driving speed information, driving direction information, and head orientation information of the target vehicle and the vehicle around [0147]); and
adding semantic information to the structured data to obtain vectorized features of the target vehicle (intention types determined based on a structure of a map in which the target vehicle is located is not fixed, and the intention can effectively improve accuracy of a behavior description of the target vehicle. In addition, target vehicle intention prediction is converted into matching between a motion status of the target vehicle and the map information, which is different from fixed-type intention classification in the existing method. Further, the lane intention and the road intention of the target vehicle assist each other, to improve generalization and accuracy of predicting the driving intention of the target vehicle. [0020]).
Regarding claim 10:
Li in view of Sun teaches all the limitations of claim 2, upon which this claim is dependent.
Li further teaches:
for each target vehicle of the plurality of target vehicles, structuring the perceived information to obtain structured data of the target vehicle (the driving information of the target vehicle or the surrounding vehicle includes information that can be sensed by the current vehicle and that affects a driving intention of the target vehicle, for example, location information, driving speed information, driving direction information, and head orientation information of the target vehicle and the vehicle around [0147]); and
adding semantic information to the structured data to obtain vectorized features of the target vehicle (intention types determined based on a structure of a map in which the target vehicle is located is not fixed, and the intention can effectively improve accuracy of a behavior description of the target vehicle. In addition, target vehicle intention prediction is converted into matching between a motion status of the target vehicle and the map information, which is different from fixed-type intention classification in the existing method. Further, the lane intention and the road intention of the target vehicle assist each other, to improve generalization and accuracy of predicting the driving intention of the target vehicle. [0020]).
Regarding claim 11:
Li in view of Sun teaches all the limitations of claim 10, upon which this claim is dependent.
Li further teaches:
performing sequential processing on the vehicle trajectory of the target vehicle and the vehicle trajectory of each of the plurality of second surrounding vehicles (driving information of the target vehicle and driving information of a surrounding vehicle of the target vehicle are obtained [0144]) to obtain sequential data comprising a historical frame and a current frame of a preset length (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 [0149]);
and converting the sequential data into a coordinate system with a current frame position of the target vehicle as an origin so that the sequential data is used as the structured data of the target vehicle (interaction feature vector that is between the target vehicle at the current moment and the i.sup.th lane and that is obtained by using the MLP network through extraction on the interaction feature between the target vehicle at the current moment and the i.sup.th lane, h.sub.li.sup.t-1 represents an interaction feature implicit vector between the target vehicle at a previous moment and the i.sup.th lane, h.sub.li.sup.t represents an interaction feature implicit vector between the target vehicle at the current moment and the i.sup.th lane, g.sub.j.sup.t represents an interaction feature between the target vehicle at the current moment …[0174]).
Regarding claim 12:
Li in view of Sun teaches all the limitations of claim 11, upon which this claim is dependent.
Li further teaches:
segmenting a lane line in the road network information according to a preset distance to obtain a plurality of lane line segments (L.sub.j is a length of a virtual lane line segment, and j=0, 1, 2, 3, 4; L.sub.5 is a distance between an end point of a previous line segment and a target projection point; and |⋅| is a vector length. [0154]);
converting endpoint coordinates of the plurality of lane line segments to the coordinate system with the current frame position of the target vehicle as the origin (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 [0151]);
for the plurality of lane line segments in the coordinate system with the current frame position of the target vehicle as the origin, selecting lane line segments from the plurality of lane line segments according to a preset third rule (the interaction feature between the target vehicle and each road may be extracted according to the following rule: extracting one or more of a location feature of the target vehicle in each second coordinate system, a distance feature between the target vehicle and an origin, a head orientation feature of the target vehicle, and a feature that a location of the target vehicle in each second coordinate system, a distance between the target vehicle and the origin, and the head orientation of the target vehicle change with a driving moment, where each second coordinate system is a rectangular coordinate system, an origin of each second coordinate system is determined based on an exit location of each road, and an x-axis direction is determined based on a driving direction of each road [0156]); and
using endpoint coordinates of the lane line segments selected (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 [0151]) as the structured data of the target vehicle (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 [0150]).
Regarding claim 13:
Li teaches:
A control device (fig. 1, computer system 112), comprising at least one processor (fig. 1, processor 113) and at least one storage apparatus storing a plurality of program codes (fig. 1, memory 114), wherein the plurality of program codes are adapted to be loaded and executed by the at least one processor to perform a method for predicting a vehicle trajectory (a vehicle driving intention prediction method [0006]), wherein the method for predicting the vehicle trajectory comprises:
obtaining vectorized features of a plurality of target vehicles (to determine an interaction feature vector between the surrounding vehicle and the target vehicle, where the interaction feature vector between the surrounding vehicle and the target vehicle represents impact of the surrounding vehicle on the target vehicle [0010]; a driving feature vector of the target vehicle relative to each of the plurality of lanes [0013]) based on perceived information (The sensor system 104 may include several sensors that can sense information about the ambient environment of the vehicle 100. [0062]) of an autonomous vehicle (vehicle 100 operates in an autonomous mode [0084]), wherein the perceived information is obtained based on data acquired by a sensor of the autonomous vehicle (the sensor system 104 may include a positioning system 122 (the positioning system may be a Global Positioning System (GPS), a BeiDou system, or another positioning system), an inertial measurement unit (IMU) 124, a radar 126, a laser rangefinder 128, and a camera 130. The sensor system 104 may further include sensors (for example, an in-vehicle air quality monitor, a fuel gauge, and an oil temperature gauge) in an internal system of the vehicle 100. Sensor data from one or more of these sensors can be used to detect an object and corresponding features (a location, a shape, a direction, a speed, and the like). Such detection and recognition are key functions of safe operation of the vehicle 100 [0063]);
obtaining a trajectory prediction result (The prediction unit 43 predicts the behavior intention and the future track of the target vehicle [0122]) based on vectorized features of each vehicle of the plurality of target vehicles (based on current map information and the target information sensed by the sensing unit [0122]), to obtain a plurality of trajectory prediction results of the plurality of target vehicles (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. This helps determine importance of the target of another surrounding vehicle of the vehicle, and filter a key target for interaction, so that the vehicle can plan the route in advance and safely pass through a complex road condition scenario. [0122]), wherein the plurality of target vehicles (extracting one or more of a location feature of each of the surrounding vehicles [0149]) comprise the autonomous vehicle (fig. 6, current vehicle) and a plurality of first surrounding vehicles (fig. 6, another vehicle), wherein the plurality of first surrounding vehicles being vehicles in a surrounding environment of the autonomous vehicle that are selected according to a preset first rule (The surrounding vehicle of the target vehicle may be understood as another vehicle that is at a specific distance from the target vehicle. The distance may be set by a user, or may be set by a skilled person, or may be related to a sensing distance of a sensor of the current vehicle. [0145]);
Controlling the autonomous vehicle based on the trajectory prediction result obtained (the application 141 may also be a program for controlling the autonomous vehicle to avoid collision with another vehicle and safely pass through an intersection [0102]).
Sun also teaches:
obtaining vectorized features of a plurality of target vehicles based on perceived information (the one or more sensors 102 installed in the target vehicle 100 captures velocity and position information of the nearby vehicles 200, as well as the velocity and position information of the target vehicle 100. The position and velocity information may be captured over multiple time steps, and used by the perception module 108 for determining a trajectory 202 of the target vehicle 100 and the nearby vehicles 200 [0042]) of an autonomous vehicle (A self-driving vehicle (autonomous vehicle) [0003]), wherein the perceived information is obtained based on data acquired by a sensor of the autonomous vehicle (the one or more sensors 102 of the target vehicle 100 capture and provide sensor data of nearby objects [0041]);
a trajectory prediction result (output a future trajectory predicted for the target vehicle 100 [0051]) based on the encoded vectorized features of each target vehicle of the plurality of target vehicles (The trajectory encoder 208 may include one or more be neural networks (e.g. LSTM, GRU, or the like) trained to extract and encode features of the trajectories for the target vehicle 100 and nearby vehicles 200. In one embodiment, the trajectory encoder 208 includes a first neural network for encoding position histories of an agent into a first vector, and a second neural network for encoding velocity histories of the agent in a second vector. [0045]), to obtain a plurality of trajectory prediction results of the plurality of target vehicles (The interaction network module 214 may be configured to model the interactions between the various agents 100, 200. In some situations, the future trajectory of the target vehicle may not only be influenced by the lanes that are subject to the driver's attention, but may also be influenced by interactions with the other vehicles 200 [0049]), wherein the plurality of target vehicles comprise the autonomous vehicle (fig. 2, target vehicle 100) and a plurality of first surrounding vehicles (fig. 2, other vehicles 200a-200c),
controlling the autonomous vehicle based on the trajectory prediction result obtained (the planning module 111 of the target vehicle may consider the various trajectories in planning a path for the target vehicle [0066]).
Li does not explicitly teach, however Sun teaches:
inputting, into an encoder module (fig. 2, trajectory encoder 208) of an encoder-decoder network model (fig. 2, prediction module 110), the vectorized features of the plurality of target vehicles to obtain encoded features of each target vehicle of the plurality of target vehicles (The trajectory encoder 208 may include one or more be neural networks (e.g. LSTM, GRU, or the like) trained to extract and encode features of the trajectories for the target vehicle 100 and nearby vehicles 200 [0045]);
obtaining, by a decoder model (fig. 2, decoder 216) of the encoder-decoder network model (the decoder module 216 is a standard LSTM network or any other recurrent neural network. The decoder module 216 may be configured to receive the trajectory feature of the target vehicle 100, the interaction features generated by the interaction network module 214, and the feature of the target lane generated by the lane attention module 212 [0051]),
It would have been obvious to one of ordinary skill in the art at the time of the effective filing date of the claimed invention to have modified Li to include the teachings as taught by Sun with a reasonable expectation of success. Both arts are in the same field of endeavor of autonomous vehicle control. Sun teaches the benefits of “A self-driving vehicle (autonomous vehicle) engages in various tasks to safely maneuver itself in a current environment. One of such tasks is predicting its future trajectory as well as the trajectory of surrounding vehicles. Trajectory prediction, however, may often be challenging due to the inherent uncertainty of the future, and the complexity of the environment navigated by the vehicle. Trajectory prediction may also be challenging as the driver's intention often determines trajectory, and intention may be hard to estimate. It is desirable, therefore, to have a system and method that tackles the challenging task of trajectory prediction for autonomous vehicles [Sun, 0003]”.
Regarding claim 14:
Li in view of Sun teaches all the limitations of claim 13, upon which this claim is dependent.
Li further teaches:
wherein the obtaining vectorized features of the plurality of target vehicles (a driving feature vector of the target vehicle relative to each of the plurality of lanes [0013]) further comprises:
obtaining, based on the perceived information of the autonomous vehicle (the processor 113 may predict a driving track of another vehicle based on a surrounding road condition and another vehicle condition that are detected by the sensor 153 [0104]), self- vehicle trajectory vectorized features (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 [0122]), surrounding trajectory vectorized features (the sensor 153 may detect an animal, a vehicle, an obstacle, or cross walk [0103]), and road network vectorized features of each target vehicle of the plurality of target vehicles (a driving feature implicit vector of the target vehicle relative to each of the plurality of lanes [0010]);
wherein the surrounding trajectory vectorized features are trajectory vectorized features of a plurality of second surrounding vehicles of a target vehicle of the plurality of target vehicles (a driving feature vector of each of the surrounding vehicles relative to the target vehicle [0012]), the plurality of second surrounding vehicles being vehicles in a surrounding environment of the target vehicle that are selected according to a preset second rule (The surrounding vehicle of the target vehicle may be understood as another vehicle that is at a specific distance from the target vehicle. The distance may be set by a user, or may be set by a skilled person, or may be related to a sensing distance of a sensor of the current vehicle. [0145]).
Regarding claim 15:
Li in view of Sun teaches all the limitations of claim 14, upon which this claim is dependent.
Li further teaches:
wherein the obtaining the trajectory prediction result (The prediction unit 43 predicts the behavior intention and the future track of the target vehicle [0122]) based on the encoded features of each target vehicle of the plurality of target vehicles (based on current map information and the target information sensed by the sensing unit [0122]) further comprises:
obtaining, for each target vehicle of the plurality of target vehicles, encoded (The target fusion unit 42 processes the environment information around the vehicle sensed by the sensing unit 41, and outputs obstacle target information [0122]) features of the target vehicle (obtaining driving information of the target vehicle [0007]) based on the self-vehicle trajectory vectorized features (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 [0122]), the surrounding trajectory vectorized features (the lane intention of the target vehicle based on the driving feature of the surrounding vehicle relative to the target vehicle [0007]), and the road network vectorized features of the target vehicle (the driving feature of the target vehicle relative to each of the plurality of roads [0008]);
and obtaining the trajectory prediction result of the target vehicle based on the encoded features of the target vehicle (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. [0122]).
Sun also further teaches:
wherein obtaining the trajectory prediction result (predicting a future trajectory of a target vehicle [0038]) based on the encoded features of each target vehicle of the plurality of target vehicles (The lane feature vectors for the lanes generated by the lane encoder 210, and the trajectory feature vector for the target vehicle 100 that is generated by the trajectory encoder 208, are used by the lane attention module 212 for identifying one or more target lanes that are predicted to be the goal of the target vehicle in the upcoming future (e.g. in the next 3 seconds) [0047]) further comprises:
obtaining, for each target vehicle of the plurality of target vehicles, encoded () features of the target vehicle (a trajectory 202 of the target vehicle 100 and the nearby vehicles 200 [0042]) based on the self-vehicle trajectory vectorized features (a trajectory 202 of the target vehicle 100 [0042]), the surrounding trajectory vectorized features (extracting lane features based on the set of ordered lane-points 204 of the lane segments. The final output of the convolutional neural network may be pooled and stored in a fixed size feature vector for each lane [0046]), and the road network vectorized features of the target vehicle (the perception module 108 is further configured to obtain a local map including a set of ordered lane-points (e.g. geographic x, y coordinates) 204 of the center of the lanes within a threshold vicinity/distance 206 of the target vehicle 100 [0043]).
Regarding claim 16:
Li in view of Sun teaches all the limitations of claim 15, upon which this claim is dependent.
Li further teaches:
wherein obtaining encoded features of the target vehicle based on the self-vehicle trajectory vectorized features (The target fusion unit 42 processes the environment information around the vehicle sensed by the sensing unit 41, and outputs obstacle target information [0122]), the surrounding trajectory vectorized features (the lane intention of the target vehicle based on the driving feature of the surrounding vehicle relative to the target vehicle [0007]), and the road network vectorized features of the target vehicle (the driving feature of the target vehicle relative to each of the plurality of roads [0008]) further comprises:
encoding the self-vehicle trajectory vectorized features, the surrounding trajectory vectorized features, and the road network vectorized features (The target fusion unit 42 processes the environment information around the vehicle sensed by the sensing unit 41, and outputs obstacle target information [0122]) to obtain self-vehicle trajectory encoded features, surrounding trajectory encoded features, and road network encoded features (The target fusion unit 42 processes the environment information around the vehicle sensed by the sensing unit 41, and outputs obstacle target information [0122]), respectively;
performing feature interaction (an interaction feature between the target vehicle and another vehicle, may be determined based on the driving information of the target vehicle and the driving information of the surrounding vehicle of the target vehicle [0148]) on the self-vehicle trajectory encoded features, the surrounding trajectory encoded features, and the road network encoded (The target fusion unit 42 processes the environment information around the vehicle sensed by the sensing unit 41, and outputs obstacle target information [0122]) features to obtain self-vehicle trajectory interaction features (the interaction feature between the target vehicle and the other vehicle [0157]), surrounding trajectory interaction features (the interaction feature between the target vehicle and each road [0157]), and environment interaction features (the interaction feature between the target vehicle and each lane [0157]), respectively ; and
performing feature fusion (The target fusion unit 42 processes the environment information around the vehicle sensed by the sensing unit 41 [0122]) on the self-vehicle trajectory interaction features, the surrounding trajectory interaction features, and the environment interaction features to obtain the encoded features of the target vehicle (The driving feature vector of each of the other vehicles relative to the target vehicle is input into the interaction feature vector prediction network, to obtain an interaction feature vector between the another vehicle and the target vehicle. [0161]).
Regarding claim 20:
Li in view of Sun teaches all the limitations of claim 13, upon which this claim is dependent.
Li further teaches:
A vehicle (fig. 1, vehicle 100), comprising the control device of claim 13 (fig. 1, controller 112).
Claim(s) 5-6 and 17-18 is/are rejected under 35 U.S.C. 103 as being unpatentable over Li et. al. (US 2023/0399023), herein Li in view of Sun et. al. (US 2021/0174668), herein Sun in further view of Kuderer et. al. (Learning Driving Styles…)(NPL), herein Kuderer.
Regarding claim 5:
Li in view of Sun teaches all the limitations of claim 4, upon which this claim is dependent.
Li further teaches:
wherein performing feature interaction on the self-vehicle trajectory encoded features, the surrounding trajectory encoded features, and the road network encoded features (an interaction feature between the target vehicle and another vehicle, may be determined based on the driving information of the target vehicle and the driving information of the surrounding vehicle of the target vehicle [0148]) further comprises:
using, for the self-vehicle trajectory encoded features, [an all-ones vector] as environment information for feature interaction to obtain the self-vehicle trajectory interaction features (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. [0150]);
using, for the surrounding trajectory encoded features, the self-vehicle trajectory encoded features as environment information for feature interaction to obtain the surrounding trajectory interaction features (a driving feature of the target vehicle relative to each road, namely, an interaction feature between the target vehicle and each road, is determined based on the driving information of the target vehicle and the road layer information [0155]); and
using, for the road network encoded features, fusion features of the self-vehicle trajectory interaction features and the surrounding trajectory interaction features as environment information for feature interaction to obtain road network interaction features (a road intention of the target vehicle and a lane intention of the target vehicle are determined based on the interaction feature between the target vehicle and the another vehicle, the interaction feature between the target vehicle and each lane, and the interaction feature between the target vehicle and each road. [0158]).
Li in view of Sun does not explicitly teach, however Kuderer teaches:
an all-ones vector (The initial guess for the feature weights θ was an all-ones vector [page 2645])
It would have been obvious to one of ordinary skill in the art at the time of the effective filing date of the claimed invention to have modified Li in view of Sun to include the teachings as taught by Kuderer with a reasonable expectation of success. Both arts are in the same field of endeavor of autonomous vehicle control. Kuderer teaches the benefit of “a learning
from demonstration approach that allows the user to simply demonstrate the desired style by driving the car manually. We model the individual style in terms of a cost function and use
feature-based inverse reinforcement learning to find the model parameters that fit the observed style best. Once the model has been learned, it can be used to efficiently compute trajectories for the vehicle in autonomous mode. We show that our approach is capable of learning cost functions and reproducing different driving styles using data from real drivers. [Kuderer, abstract]”.
Regarding claim 6:
Li in view of Sun and Kuderer teaches all the limitations of claim 5, upon which this claim is dependent.
Li further teaches:
performing feature fusion on the self-vehicle trajectory interaction features and the surrounding trajectory interaction features to obtain the fusion features (An algorithm of predicting the road intention of the target vehicle is as follows… h.sub.gj.sup.t represents an interaction feature implicit vector between the target vehicle at a current moment and a j.sup.th road, h.sub.li.sup.t represents an interaction feature implicit vector between the target vehicle and an i.sup.th lane, α.sub.ji represents a lane intention of the target vehicle corresponding to the i.sup.th lane, P.sub.j.sup.t represents a fusion vector of an interaction feature implicit vector between the i.sup.th lane and the target vehicle corresponding to all lanes associated with the j.sup.th road, and β.sub.j represents a probability that the target vehicle drives away from the intersection from the j.sup.th road, namely, the road intention of the target vehicle corresponding to the j.sup.th road, that is obtained after h.sub.gj.sup.t, and P.sub.j.sup.t are input into the road intention prediction subnetwork [0183]).
Regarding claim 17:
Li in view of Sun teaches all the limitations of claim 16, upon which this claim is dependent.
Li further teaches:
wherein performing feature interaction on the self-vehicle trajectory encoded features, the surrounding trajectory encoded features, and the road network encoded features (an interaction feature between the target vehicle and another vehicle, may be determined based on the driving information of the target vehicle and the driving information of the surrounding vehicle of the target vehicle [0148]) further comprises:
using, for the self-vehicle trajectory encoded features, [an all-ones vector] as environment information for feature interaction to obtain the self-vehicle trajectory interaction features (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. [0150]);
using, for the surrounding trajectory encoded features, the self-vehicle trajectory encoded features as environment information for feature interaction to obtain the surrounding trajectory interaction features (a driving feature of the target vehicle relative to each road, namely, an interaction feature between the target vehicle and each road, is determined based on the driving information of the target vehicle and the road layer information [0155]); and
using, for the road network encoded features, fusion features of the self-vehicle trajectory interaction features and the surrounding trajectory interaction features as environment information for feature interaction to obtain road network interaction features (a road intention of the target vehicle and a lane intention of the target vehicle are determined based on the interaction feature between the target vehicle and the another vehicle, the interaction feature between the target vehicle and each lane, and the interaction feature between the target vehicle and each road. [0158]).
Li in view of Sun does not explicitly teach, however Kuderer teaches:
an all-ones vector (The initial guess for the feature weights θ was an all-ones vector [page 2645])
It would have been obvious to one of ordinary skill in the art at the time of the effective filing date of the claimed invention to have modified Li in view of Sun to include the teachings as taught by Kuderer with a reasonable expectation of success. Both arts are in the same field of endeavor of autonomous vehicle control. Kuderer teaches the benefit of “a learning
from demonstration approach that allows the user to simply demonstrate the desired style by driving the car manually. We model the individual style in terms of a cost function and use
feature-based inverse reinforcement learning to find the model parameters that fit the observed style best. Once the model has been learned, it can be used to efficiently compute trajectories for the vehicle in autonomous mode. We show that our approach is capable of learning cost functions and reproducing different driving styles using data from real drivers. [Kuderer, abstract]”.
Regarding claim 18:
Li in view of Sun and Kuderer teaches all the limitations of claim 17, upon which this claim is dependent.
Li further teaches:
performing feature fusion on the self-vehicle trajectory interaction features and the surrounding trajectory interaction features to obtain the fusion features (An algorithm of predicting the road intention of the target vehicle is as follows… h.sub.gj.sup.t represents an interaction feature implicit vector between the target vehicle at a current moment and a j.sup.th road, h.sub.li.sup.t represents an interaction feature implicit vector between the target vehicle and an i.sup.th lane, α.sub.ji represents a lane intention of the target vehicle corresponding to the i.sup.th lane, P.sub.j.sup.t represents a fusion vector of an interaction feature implicit vector between the i.sup.th lane and the target vehicle corresponding to all lanes associated with the j.sup.th road, and β.sub.j represents a probability that the target vehicle drives away from the intersection from the j.sup.th road, namely, the road intention of the target vehicle corresponding to the j.sup.th road, that is obtained after h.sub.gj.sup.t, and P.sub.j.sup.t are input into the road intention prediction subnetwork [0183]).
Claim(s) 7-8 and 19 is/are rejected under 35 U.S.C. 103 as being unpatentable over Li et. al. (US 2023/0399023), herein Li in view of Sun et. al. (US 2021/0174668), herein Sun in further view of Sahin et. al. (US 2025/0330286), herein Sahin.
Regarding claim 7:
Li in view of Sun teaches all the limitations of claim 3, upon which this claim is dependent.
Li further teaches:
obtaining multimodal features of the target vehicle based on the encoded features of the target vehicle (This application provides a vehicle driving intention prediction method. A road intention and a lane intention of a target vehicle are first determined based on [0129]); and
obtaining the trajectory prediction result of the target vehicle based on the multimodal features (a driving feature of a surrounding vehicle relative to the target vehicle, a driving feature of the target vehicle relative to a road, and a driving feature of the target vehicle relative to a lane, and then a driving intention of the target vehicle is determined based on the road intention and the lane intention of the target vehicle. The driving intention of the target vehicle is determined by predicting a multi-level intention (namely, the lane intention and the road intention) of the target vehicle [0129]).
Li in view of Sun does not explicitly teach, however Sahin teaches:
using pre-learned anchor features as environment information (the RSU/anchor 102A may identify one or more additional RSU(s) along the future predicted trajectory of the detected UE 104A and activate one or more additional SL PRS transmissions [0057] [page 5 of provisional]), and
It would have been obvious to one of ordinary skill in the art at the time of the effective filing date of the claimed invention to have modified Li in view of Sun to include the teachings as taught by Sahin with a reasonable expectation of success. Both arts are in the same field of endeavor of autonomous vehicle control. Sahin teaches the benefits of “the network or anchors can proactively configure SL PRS transmissions to be activated. In one example, the proactive configuration can be based on UE density, directivity, etc. or any other information carried via the SL transmissions of the detected one or more UE(s). In another example, the proactive configuration can be based on receipt of messages explicitly indicating the positioning request of the one or more UE(s). The proactive configuration will illustratively include additional anchors (one or more) along the future predicted trajectory of the detected UEs [Sahin, 0027]”.
Regarding claim 8:
Li in view of Sun teaches all the limitations of claim 4, upon which this claim is dependent.
Li further teaches:
obtaining multimodal features of the target vehicle based on the encoded features of the target vehicle (This application provides a vehicle driving intention prediction method. A road intention and a lane intention of a target vehicle are first determined based on [0129]); and
obtaining the trajectory prediction result of the target vehicle based on the multimodal features (a driving feature of a surrounding vehicle relative to the target vehicle, a driving feature of the target vehicle relative to a road, and a driving feature of the target vehicle relative to a lane, and then a driving intention of the target vehicle is determined based on the road intention and the lane intention of the target vehicle. The driving intention of the target vehicle is determined by predicting a multi-level intention (namely, the lane intention and the road intention) of the target vehicle [0129]).
Li in view of Sun does not explicitly teach, however Sahin teaches:
using pre-learned anchor features as environment information (the RSU/anchor 102A may identify one or more additional RSU(s) along the future predicted trajectory of the detected UE 104A and activate one or more additional SL PRS transmissions [0057] [page 5 of provisional]), and
It would have been obvious to one of ordinary skill in the art at the time of the effective filing date of the claimed invention to have modified Li in view of Sun to include the teachings as taught by Sahin with a reasonable expectation of success. Both arts are in the same field of endeavor of autonomous vehicle control. Sahin teaches the benefits of “the network or anchors can proactively configure SL PRS transmissions to be activated. In one example, the proactive configuration can be based on UE density, directivity, etc. or any other information carried via the SL transmissions of the detected one or more UE(s). In another example, the proactive configuration can be based on receipt of messages explicitly indicating the positioning request of the one or more UE(s). The proactive configuration will illustratively include additional anchors (one or more) along the future predicted trajectory of the detected UEs [Sahin, 0027]”.
Regarding claim 19:
Li in view of Sun teaches all the limitations of claim 15, upon which this claim is dependent.
Li further teaches:
obtaining multimodal features of the target vehicle based on the encoded features of the target vehicle (This application provides a vehicle driving intention prediction method. A road intention and a lane intention of a target vehicle are first determined based on [0129]); and
obtaining the trajectory prediction result of the target vehicle based on the multimodal features (a driving feature of a surrounding vehicle relative to the target vehicle, a driving feature of the target vehicle relative to a road, and a driving feature of the target vehicle relative to a lane, and then a driving intention of the target vehicle is determined based on the road intention and the lane intention of the target vehicle. The driving intention of the target vehicle is determined by predicting a multi-level intention (namely, the lane intention and the road intention) of the target vehicle [0129]).
Li in view of Sun does not explicitly teach, however Sahin teaches:
using pre-learned anchor features as environment information (the RSU/anchor 102A may identify one or more additional RSU(s) along the future predicted trajectory of the detected UE 104A and activate one or more additional SL PRS transmissions [0057] [page 5 of provisional]), and
It would have been obvious to one of ordinary skill in the art at the time of the effective filing date of the claimed invention to have modified Li in view of Sun to include the teachings as taught by Sahin with a reasonable expectation of success. Both arts are in the same field of endeavor of autonomous vehicle control. Sahin teaches the benefits of “the network or anchors can proactively configure SL PRS transmissions to be activated. In one example, the proactive configuration can be based on UE density, directivity, etc. or any other information carried via the SL transmissions of the detected one or more UE(s). In another example, the proactive configuration can be based on receipt of messages explicitly indicating the positioning request of the one or more UE(s). The proactive configuration will illustratively include additional anchors (one or more) along the future predicted trajectory of the detected UEs [Sahin, 0027]”.
Conclusion
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure.
Fan (US 2025/0137796) discloses a trajectory prediction method and an apparatus therefor, a medium, a program product, and an electronic device. An example trajectory prediction method includes: obtaining historical trajectory information of a target vehicle and an associated vehicle; predicting location distribution information of the target vehicle and the associated vehicle based on the historical trajectory information and map information; determining an interaction feature between the target vehicle and the associated vehicle based on the location distribution information; and determining a traveling trajectory of the target vehicle based on the interaction feature, the location distribution information of the target vehicle, and the map information.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to Scott R Jagolinzer whose telephone number is (571)272-4180. The examiner can normally be reached M-Th 8AM - 4PM Eastern.
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, Christian Chace can be reached at (571)272-4190. 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.
Scott R. Jagolinzer
Examiner
Art Unit 3665
/S.R.J./Examiner, Art Unit 3665 /CHRISTIAN CHACE/Supervisory Patent Examiner, Art Unit 3665
Prosecution Insights