Notice of Pre-AIA or AIA Status
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
Response to Amendment
The previous 35 U.S.C. 112(b) rejections are withdrawn due to Applicant’s amendments.
Response to Arguments
Applicant’s arguments filed 01-12-2026 on pages 7-11 of Remarks regarding the rejection under 35 U.S.C. 103 with respect to claims 1-22 have been fully considered but are moot. New references Lehre and Shen have been incorporated below to teach the newly presented limitations.
Claim Rejections - 35 USC § 112(a) – Written Description
The following is a quotation of the first paragraph of 35 U.S.C. 112(a):
(a) IN GENERAL.—The specification shall contain a written description of the invention, and of the manner and process of making and using it, in such full, clear, concise, and exact terms as to enable any person skilled in the art to which it pertains, or with which it is most nearly connected, to make and use the same, and shall set forth the best mode contemplated by the inventor or joint inventor of carrying out the invention.
The following is a quotation of the first paragraph of pre-AIA 35 U.S.C. 112:
The specification shall contain a written description of the invention, and of the manner and process of making and using it, in such full, clear, concise, and exact terms as to enable any person skilled in the art to which it pertains, or with which it is most nearly connected, to make and use the same, and shall set forth the best mode contemplated by the inventor of carrying out his invention.
Claims 16 and 17 are rejected under 35 U.S.C. 112(a) or 35 U.S.C. 112 (pre-AIA ), first paragraph, as failing to comply with the written description requirement. The claim(s) contains subject matter which was not described in the specification in such a way as to reasonably convey to one skilled in the relevant art that the inventor or a joint inventor, or for applications subject to pre-AIA 35 U.S.C. 112, the inventor(s), at the time the application was filed, had possession of the claimed invention.
As per MPEP 2161.01, a computer-implemented functional claim limitation may lack adequate written description when the claims define the invention in functional language specifying a desired result but the specification does not sufficiently describe how the function is performed or the result is achieved. In other words, the algorithm or steps/procedure taken to perform the function must be described with sufficient detail to that one of ordinary skill in the art would understand how the inventor intended the function to be performed. Moreover, the Federal Circuit has explained that a specification cannot always support expansive claim language and satisfy the requirements of 35 U.S.C. 112 “merely by clearly describing one embodiment of the thing claimed.” LizardTech v. Earth Resource Mapping, Inc., 424 F.3d 1336, 1346, 76 USPQ2d 1731, 1733 (Fed. Cir. 2005).
If it is the position of Applicant that any of the functions identified below are so well known in the art that they need not be described in the specification, Applicant should state this clearly on the record. This will be taken as an admission when considering prior art rejections. The issue is whether a person skilled in the art would understand the inventor to have invented, and been in possession of, the invention as broadly claimed.
Dependent claims 16 and 17 recite “determining by the processor to operate during a further period of time, in a further configuration that is slower than the given mode of object detections.” The scope of the claim encompasses all possible ways of operating in a certain configuration that is slower than other configurations of the processor. In contrast, the specification, see e.g., published [0042], describes at best a configuration that its main purpose is to operate quickly and allow a high frame rate. This limited disclosure does not support the expansive claim language. The specification does not mention how a processor determines a configuration to be slower than another configuration in order to operate.
Claim Rejections - 35 USC § 112(b)
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 16 and 17 are rejected under 35 U.S.C. 112(b) or 35 U.S.C. 112 (pre-AIA ), second paragraph, as failing to set forth the subject matter which the inventor or a joint inventor, or for applications subject to pre-AIA 35 U.S.C. 112, the applicant regards as the invention.
Dependent claims 16 and 17 recite “determining by the processor to operate during a further period of time, in a further configuration that is slower than the given mode of object detections.” It is unclear what exactly is meant by “slower than” as it is not clear what the other configurations are and whether the certain configurations would be slower.
Claim Rejections - 35 USC § 103
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
Claims 1, 5, 6, 7, 8, 10, 11, 13, 15, 16, 18, 20, 21 and 22 are rejected under 35 U.S.C. 103 as being unpatentable over Mansour et al. (US11200438B2); hereinafter Mansour in view of Stein et al. (US 20190325595 A1); hereinafter Stein in view of Shen et al. (US 20200175401 A1); hereinafter Shen and in further view of Lehre et al. (US 20090204289 A1); hereinafter Lehre
Claim 1 is rejected over Mansour, Stein, Shen and Lehre.
Regarding claim 1, Mansour teaches a method for determining driving related features of a vehicle, the method comprises:
obtaining, by the processor, sensed information during at least a part of the time interval; (Mansour [col 3, lines 35-37]: “The HCNN system 10 may further receive additional input images over a predetermined period of time, for example once every 30 ms as the host vehicle 12 travels along the highway 14.”;)
selecting, by the processor, based on the processing and using the context information, a selected sub-set of narrow artificial intelligence (AI) agents; (Mansour [col. 6, lines 11-30]: “The method 100 is configured to train any number of tasks 1, 2 . . . n where n is a whole number greater than 1. The tasks 1, 2 . . . n may, for example, correspond to the first sub-network 44, the second sub-network 46, up to the n sub-network 47, respectively. Thus, these tasks may include lane detection and object detection, as described above.”; col. 5, lines 59-65; and “For each of the batches 1, 2 . . . x, a set of annotated images are assigned. These annotated images correspond to the task 1. For example, where task 1 is lane detection, then the annotated images assigned to batches 1, 2 . . . x include annotated lanes that train the HCNN on lane detection.”; Note: The sub-networks are narrow AI agents.)
Mansour does not teach receiving context information defining a distance range of interest associated with the given time interval and pertaining to a driving of the vehicle at the given time interval, wherein the context information is determined at a point of time preceding the given time interval;
processing, by the processor, the context information in accordance with the sensed information;
calculating, by the processor, one or more driving related features, wherein the calculating comprises applying the selected sub-set of narrow AI agents to process only pixels belonging to a first portion of the sensed information that corresponds to the distance range of interest, and ignoring a second portion of the sensed information that corresponds to one or more distances ranges that are outside of the distance range of interest; the ignoring comprising avoiding processing pixels of the second portion
wherein the one or more driving related features are for use in a response selected out of autonomously driving the vehicle based on the one or more driving related features, or performing a driver assistance system (ADAS) operation based on the one or more driving related features; and
determining, based on the sensed information, subsequent context information defining a distance range of interest associated with a next time interval that follows the given time interval and pertaining to a driving of the vehicle at the next time interval, and
providing the one or more driving related features for autonomous driving of the vehicle during the next time interval.
However, Stein teaches receiving context information defining a distance range of interest associated with the given time interval and pertaining to a driving of the vehicle at the given time interval, wherein the context information is determined at a point of time preceding the given time interval; (Stein [0164]: “Operation 1906 may further assess whether any detected pedestrians are within the vicinity (e.g., within “splash range”) of a detected puddle. The pedestrian assessment may be used to computationally determine the degree of preference for taking evasive action in the control of the autonomous vehicle to avoid splashing the pedestrian. In an example, the splash range may be predefined (context information), such as a certain distance from the center or edges of a puddle. In an example, the splash range may be estimated, such as based on the size of the puddle, the speed of the vehicle, etc.”; and [0382]: “processing the sequence of images to detect a puddle on the road; determining any presence of a pedestrian in a vicinity of the puddle;)
processing the context information in accordance with the sensed information; (Stein [0382]: “processing the sequence of images to detect a puddle on the road; determining any presence of a pedestrian in a vicinity of the puddle;)
calculating, by the processor, one or more driving related features, wherein the calculating comprises applying the selected sub-set of narrow AI agents to process only pixels belonging to a first portion of the sensed information that corresponds to the distance range of interest, and ignoring a second portion of the sensed information that corresponds to one or more distances ranges that are outside of the distance range of interest; the ignoring comprising avoiding processing pixels of the second portion (Stein [0378]: “In Example 122, the subject matter of Example 121 includes, wherein the at least one processing device is configured to: determine from the plurality of images if a target is located within the splash zone of the vehicle passing through the puddle (first portion of the sensed information); and control the vehicle to perform a navigational maneuver to modify the splash zone of the vehicle such that a new splash zone of the vehicle does not include the target.”; and [0382]: “Example 126 is a method for controlling an autonomous vehicle traveling along a road, the method being carried out by computing platform, and comprising: storing a sequence of images representing at least one field of view from a perspective of the vehicle that includes, a portion of the road; processing the sequence of images to detect a puddle on the road; determining any presence of a pedestrian in a vicinity of the puddle; and determining a driving response solution in response to detection of the puddle, wherein the driving response solution is based on whether the presence of the pedestrian was detected.”; Note: A splash zone is a calculated region of interest where whatever is not in the splash zone is ignored and if a target is detected within the splash zone, those pixels are processed and evasion action from the vehicle is taken. No action is taken if the target is outside of the splash zone, therefore the target that is outside of the splash zone is the second portion of the sensed information that corresponds to the distance ranges that are outside of the distance range of interest that is ignored.)
wherein the one or more driving related features are for use in a response selected out of autonomously driving the vehicle based on the one or more driving related features, or performing a driver assistance system (ADAS) operation based on the one or more driving related features; and (Stein [0121]: “FIG. 15 is a flow diagram illustrating an example of a method 1500 for real-time measurement of vertical contour of a road while an autonomous vehicle is moving along the road, according to an embodiment. The operations of the method 1500 are performed by computational hardware, such as that described above or below (e.g., processing circuitry).”; and [0155]: “At operation 1714, the available driving response solutions are assessed based on the current situational scenario. The driving response solutions (driving related feature) in this example are specific to responding to the detected presence of the puddle. The method 2000 provides an example of operation 1714. At operation 1716, a driving response solution is selected based on assessment of various driving response options. Notably, the driving response solution may forgo taking any evasive or other action.”;)
determining, based on the sensed information, subsequent context information defining a distance range of interest associated with a next time interval that follows the given time interval and pertaining to a driving of the vehicle at the next time interval, and (Stein [0379]: “In Example 123, the subject matter of Example 122 includes, wherein the navigational maneuver is at least one of slowing the vehicle, performing an intra-lane swerve such that a new path of the vehicle is far enough from the target to prevent the target from being struck by a splash, or perform an intra-lane swerve such that the new path of the vehicle no longer passes through the puddle.”; and [0382]: “determining a driving response solution in response to detection of the puddle, wherein the driving response solution is based on whether the presence of the pedestrian was detected.”;)
providing the one or more driving related features for autonomous driving of the vehicle during the next time interval. (Stein [0121]: “FIG. 15 is a flow diagram illustrating an example of a method 1500 for real-time measurement of vertical contour of a road while an autonomous vehicle is moving along the road, according to an embodiment. The operations of the method 1500 are performed by computational hardware, such as that described above or below (e.g., processing circuitry).”; and [0155]: “At operation 1714, the available driving response solutions are assessed based on the current situational scenario. The driving response solutions (driving related feature) in this example are specific to responding to the detected presence of the puddle. The method 2000 provides an example of operation 1714. At operation 1716, a driving response solution is selected based on assessment of various driving response options. Notably, the driving response solution may forgo taking any evasive or other action.”;)
It would have been obvious before the effective filing date to combine the AI agents of Mansour and the calculation of driving related features of Stein for effective maneuvering of autonomous vehicles (Stein, paragraph [0162]). Mansour and Stein are analogous art because they both concern autonomous vehicles.
Mansour does not teach sequentially operating at different modes of object detection during different time intervals of a sub-period;
wherein an operating at a given time interval of the different time interval comprises:
determining by a processor that is a hardware neural network processor that comprises one or more integrated circuits, to operate during the time interval,
However, Shen teaches sequentially operating at different modes of object detection during different time intervals of a sub-period; (Shen [0044]: “In some embodiments, the outputs of the first image processing engine, second image processing engine, or both are cached. If the image processing network 102 cannot run two neural networks at the same time, for example, then execution of the first image processing engine will have to be paused by system 100 while the second image processing engine runs. After the second image processing engine is finished running, system 100 will have to re-execute. In other embodiments, rather than caching the outputs of the image processing engines, system 100 executes the first and second image processing engine serially, or executes the first and second image processing engine in parallel.”)
wherein an operating at a given time interval of the different time interval comprises: (Shen [0052]: “FIG. 4 is an illustration of an example 400 of object detection using a first detector and a second detector. Detector 408 in example 400 produces outputs at 80 ms latency, and is paired with faster detector 406, which produces outputs at 15 ms latency. The camera produces output frames at 33 ms. In some embodiments, both detector 406 and detector 408 are deep learning neural networks. In some embodiments, detector 408 employs a high compute, low resolution neural network, while detector 406 employs a low compute, low resolution neural network.”)
determining by a processor that is a hardware neural network processor that comprises one or more integrated circuits, to operate during the time interval, (Shen [0035]: “In variants where the processing hardware's computation resources are limited (e.g., microcontrollers, ASICS, etc.), the system can automatically cluster layers of the first and/or second image processing engines into blocks (e.g., in variants wherein the first and/or second image processing engines include neural networks), such that the image processing engines can be interrupted when more urgent functions need to be executed.”)
It would have been obvious before the effective filing date to combine the AI agents of Mansour with the different frequencies of Shen to improve image processing accuracy for object detection (Shen [0012]). Mansour and Shen are analogous art because they both concern object detection in autonomous vehicles.
Mansour does not teach in a certain mode of object detection of the different modes of object detection
wherein the different modes of object detection are different combinations of at least two of (a) one or more types of objects to be detected, (b) distance ranges of interest, and (c) width of field of view;
However, Lehre teaches in a certain mode of object detection of the different modes of object detection (Lehre [0004]: “This object is attained by providing a coordinator which, on the one hand, makes available the ranges by the object detection sensor for its respective operating modes in which objects to be detected statistically are to be expected, and on the other hand, informs the coordinator of the driver assistance functions in which sensor recording range parts the objects are to be detected that are required for their functionality. This can be solved by the coordinator in such a way that it correlates the respective distribution density functions with one another, and thus determines the suitable operating mode of the object detection sensor for each driver assistance function, and, corresponding to the instantaneously activated driver assistance function, appropriately switches over the object detection sensor between its operating modes.”)
wherein the different modes of object detection are different combinations of at least two of (a) one or more types of objects to be detected, (b) distance ranges of interest, and (c) width of field of view; (Lehre [0008]: “it is advantageous that the detection probability density function states in which sub-range, of the sensor recording range objects are able to be detected particularly well based on the selected operating mode, in that the function gives the detection probability density as a function of the clearance and/or the azimuth angle of the sensor recording range.”; [0031]: “The figure shows an object detection sensor 1 which may be developed as a radar sensor, for example, and has two operating modes, a first operating mode A having a great operating range, and a second operating mode B having a lesser operating range which, however, has a greater resolution accuracy and which ascertains the object positions of the detected objects in shorter measuring cycles than operating mode A.”; Note: Modes A and B are distance ranges of interest and azimuth is the width of field of view. Switching between modes are the different combinations.)
It would have been obvious before the effective filing date to combine the AI agents of Mansour with the mode switching of Lehre to advantageously operate between multiple modes and accuracies (Lehre [0011]). Mansour and Lehre are analogous art because they both concern object detection in autonomous vehicles.
Claim 5 is rejected over Mansour, Stein, Shen and Lehre with the incorporation of claim 1.
Regarding claim 5, Mansour does not teach wherein the context information includes information pertaining to a field of view.
However, Stein teaches wherein the context information includes information pertaining to a field of view. (Stein [0188]: “As depicted, the system includes a camera or image sensor 2112 mounted in or on vehicle. Each image sensor 2112 images a field of view, to provide image frames 2115, which are read by the image processor 2130. In an example, more than one camera 2112 may be mounted in the vehicle.”; [0188])
It would have been obvious before the effective filing date to combine the AI agents of Mansour and the calculation of driving related features of Stein for effective maneuvering of autonomous vehicles (Stein, paragraph [0162]). Mansour and Stein are analogous art because they both concern autonomous vehicles.
Claim 6 is rejected over Mansour, Stein, Shen and Lehre with the incorporation of claim 1.
Regarding claim 6, Mansour teaches wherein the sub-set consists of a single narrow AI agent. (Mansour [col 4, lines 8-15]: “As noted above the HCNN system 10 performs several parallel tasks. A first sub-network 44 performs a first task of object detection, classification, and localization for certain classes of objects (vehicles, pedestrians, traffic signs, traffic lights, and the like,) where the output from the first sub-network 44 is the list of detected objects, detected object table 32, which provides a confidence level and location information for the detected objects.”; Note: A first sub-network is a single AI agent)
Claim 7 is rejected over Mansour, Stein, Shen and Lehre with the incorporation of claim 1.
Regarding claim 7, Mansour teaches wherein the sub-set comprises two or more narrow AI agents. (Mansour [col. 5, lines 7-11]: “The HCNN system 10 combines different sub-networks such as the first sub-network 44 and the second sub-network 46 to perform multiple tasks efficiently, thereby using a smaller memory footprint (memory saving) and operating faster than running the different sub-networks separately.”;)
Claim 8 is rejected over Mansour, Stein, Shen and Lehre with the incorporation of claim 1.
Regarding claim 8, Mansour teaches wherein the applying the selected sub-set of narrow AI agents on the sensed information provides at least one narrow AI output; (Mansour [col. 3, lines 41-46]: “A first network output defining a detected object table 32 provides a list of detected objects, including object types 34 such as a car, a truck, a pedestrian, and the like, and a confidence level 36 in the accuracy of defining the object type 34. Production of the detected object table 32 requires solutions of classification and localization of the objects.”;)
and wherein the calculating further comprises processing the at least one narrow AI output to provide the one or more driving related feature (Mansour [col. 3, lines 46-51]: “A second network output defining a segmentation data set 38 provides data to the host vehicle 12 related to lane detection, lane conditions, and lane positions relative to the host vehicle 12 within the transmission window 28 of the HCNN system 10.”;)
Claim 10 is rejected over Mansour, Stein, Shen and Lehre with the incorporation of claim 1.
Regarding claim 10, Mansour teaches performing an advance driver assistance system (ADAS) operation based on the at least one driving related feature. (Mansour [col. 5, lines 43-48]: “The HCNN system 10 is described in one present example for use in a perception kit for an ADAS and autonomous vehicle vision system. The HCNN system 10 performs two tasks simultaneously, which in the example of the ADAS and autonomous vehicle vision system includes lane detection and object detection”;)
Claim 11 is rejected over is rejected over Mansour, Stein, Shen and Lehre with the incorporation of claim 1.
Regarding claim 11, Mansour teaches wherein at least one narrow AI agent of the selected sub-set of narrow AI agents is a part of a neural network. (Mansour [col 4, lines 8-15]: “As noted above the HCNN system 10 performs several parallel tasks. A first sub-network 44 performs a first task of object detection, classification, and localization for certain classes of objects (vehicles, pedestrians, traffic signs, traffic lights, and the like,) where the output from the first sub-network 44 is the list of detected objects, detected object table 32, which provides a confidence level and location information for the detected objects.”; Note: A first sub-network is a single AI agent)
Claim 13 is rejected over Mansour, Stein, Shen and Lehre.
Regarding claim 13, Mansour teaches a non-transitory computer readable medium that stores instructions for. (Mansour [col. 3, lines 54-60]: “The controller 40 is a non-generalized, electronic control device having a preprogrammed digital computer or processor 42, memory or non-transitory computer readable medium 43 used to store data such as control logic, software applications, instructions, computer code, data, lookup tables, etc., and input/output ports 45.”;)
The remainder of claim 13 is claim 1 in the form of a non-transitory computer readable medium and is rejected for the same reasons as claim 1 stated above.
Claim 15 is rejected over Mansour, Stein, Shen and Lehre with the incorporation of claim 1.
Regarding claim 15, Mansour teaches performing the ADAS operation based on the one or more driving related features. (Mansour [col. 5, lines 43-48]: “The HCNN system 10 is described in one present example for use in a perception kit for an ADAS and autonomous vehicle vision system. The HCNN system 10 performs two tasks simultaneously, which in the example of the ADAS and autonomous vehicle vision system includes lane detection and object detection”;)
Claim 16 is rejected over Mansour, Stein, Shen and Lehre with the incorporation of claim 1.
Regarding claim 16, Mansour does not teach wherein the different modes of object detection comprise a first mode of object detection that is slower than a second mode of object detection of the different modes of object detection,
the first mode of object detections, does not impose limitation on distances of pixels to be processed and is dedicated to detect certain objects;
However, Lehre teaches wherein the different modes of object detection comprise a first mode of object detection that is slower than a second mode of object detection of the different modes of object detection, (Lehre [0031]: “The figure shows an object detection sensor 1 which may be developed as a radar sensor, for example, and has two operating modes, a first operating mode A having a great operating range, and a second operating mode B having a lesser operating range which, however, has a greater resolution accuracy and which ascertains the object positions of the detected objects in shorter measuring cycles than operating mode A.”)
the first mode of object detections, does not impose limitation on distances of pixels to be processed and is dedicated to detect certain objects; (Lehre [0034]: “Objects that travel ahead of one's own vehicle at a distance greater than target object clearance dZO are of subordinate importance for ACC function 3, “; and [0036]: ”This function illustrates at which distance ranges in a respective operating mode one should count on objects to be detected first. According to this, curve 7 of remote area mode A is shaped so that it runs approximately constantly up to clearance d=dZO, and thereafter, for instance, up to the maximum operating range of the sensor at d=250 meter, decreases to zero.”)
It would have been obvious before the effective filing date to combine the AI agents of Mansour with the mode switching of Lehre to advantageously operate between multiple modes and accuracies (Lehre [0011]). Mansour and Lehre are analogous art because they both concern object detection in autonomous vehicles.
Claim 18 is rejected over Mansour, Stein, Shen and Lehre with the incorporation of claim 1.
Regarding claim 18, Mansour teaches wherein the obtaining of the sensed information comprises sensing the sensed information by one or more sensors of the vehicle and providing the sensed information to the processor. (Mansour [col 3., lines 18-31]: “The HCNN system 10 receives image data via a visual reception system 22 such as a camera, a LIDAR, or a RADAR system which collects the object attribute data, for example as a pixel image 30 shown and described in reference to FIG. 2. In this manner the object attribute data may be utilized for Advanced Driver Assist (ADAS) technology by also utilizing sensors that are in an existing centralized vision processor. The visual reception system 22 may further receive information as object imaging data defining the pedestrian 26 in an immediate vicinity of the fourth vehicle 24, and fixed objects such as bridges, guard rails, trees, highway signs, and the like that are all located within a host vehicle predefined sensing and transmission window 28 of the HCNN system 10.”;)
Claim 20 is rejected over Mansour, Stein, Shen and Lehre with the incorporation of claim 1.
Regarding claim 20, Mansour does not teach comprising selecting a configuration out of different configurations, based on a system constraint selected out of system constraints, wherein each of the different configurations is associated with a unique set of different modes of object detection.
However, Lehre teaches comprising selecting a configuration out of different configurations, based on a system constraint selected out of system constraints, wherein each of the different configurations is associated with a unique set of different modes of object detection. (Lehre [0004]: “This can be solved by the coordinator in such a way that it correlates the respective distribution density functions with one another, and thus determines the suitable operating mode of the object detection sensor for each driver assistance function, and, corresponding to the instantaneously activated driver assistance function, appropriately switches over the object detection sensor between its operating modes.”)
It would have been obvious before the effective filing date to combine the AI agents of Mansour with the mode switching of Lehre to advantageously operate between multiple modes and accuracies (Lehre [0011]). Mansour and Lehre are analogous art because they both concern object detection in autonomous vehicles.
Claim 21 is rejected over Mansour, Stein, Shen and Lehre with the incorporation of claim 1.
Regarding claim 21, Mansour does not teach wherein the system constraints comprise (a) performing object detection of one or more certain types objects,
(b) facilitating a frame rate while limiting at least one of field of view width or distance ranges of interest,
(c) operating on a wide field of view while limiting at least one of a type of detected objects or the distance ranges of interest, or
(d) providing a coverage of a distance range while limiting at least one of the type of detected objects or the distance ranges of interest.
However, Lehre teaches wherein the system constraints comprise (a) performing object detection of one or more certain types objects, (Lehre [0031]: “first driver assistance function 3 carrying out an ACC function which performs a clearance regulation on superhighway-like country roads or freeways, in order to regulate the speed of one's own vehicle in the sense of a constant clearance regulation from behind a preceding vehicle that has been recognized as target object. As a further driver assistance function 4, examined in an exemplary fashion, an automatically triggered emergency brake function is shown, which, as a function of detected objects located ahead, ascertains whether a collision of one's own vehicle with a detected object is probable and, in the case where a collision is no longer avoidable, carries out an automatically triggered and automatically executed emergency braking.”)
It would have been obvious before the effective filing date to combine the AI agents of Mansour with the mode switching of Lehre to advantageously operate between multiple modes and accuracies (Lehre [0011]). Mansour and Lehre are analogous art because they both concern object detection in autonomous vehicles.
Claim 22 is rejected over Mansour, Stein, Shen and Lehre with the incorporation of claim 1.
Regarding claim 22, Mansour does not teach one mode of object detection of the different modes of object detection operates at a higher frame rate than another object detection of the different modes of object detection.
However, Shen teaches one mode of object detection of the different modes of object detection operates at a higher frame rate than another object detection of the different modes of object detection. (Shen [0052]: “FIG. 4 is an illustration of an example 400 of object detection using a first detector and a second detector. Detector 408 in example 400 produces outputs at 80 ms latency, and is paired with faster detector 406, which produces outputs at 15 ms latency. The camera produces output frames at 33 ms. In some embodiments, both detector 406 and detector 408 are deep learning neural networks. In some embodiments, detector 408 employs a high compute, low resolution neural network, while detector 406 employs a low compute, low resolution neural network. In some embodiments, the faster detector 406 has the ability to use information from the previous frame.”)
It would have been obvious before the effective filing date to combine the AI agents of Mansour with the different frequencies of Shen to improve image processing accuracy for object detection (Shen [0012]). Mansour and Shen are analogous art because they both concern object detection in autonomous vehicles.
Claims 2, 3, 9, 12 and 14 are rejected under 35 U.S.C. 103 as being unpatentable over Mansour, Stein, Shen and Lehre as applied above and in further view of Zhang et al. (US 20220105926 A1); hereinafter Zhang
Claim 2 is rejected over Mansour, Stein, Shen, Lehre and Zhang with the incorporation of claim 1.
Regarding claim 2, Mansour does not teach wherein the context information further comprises at least one sensed information acquisition related feature that defines a manner in which sensed information is acquired by a sensing unit.
However, Zhang teaches wherein the context information further comprises at least one sensed information acquisition related feature that defines a manner in which sensed information is acquired by a sensing unit. (Zhang [0063]: “an illumination sensor may collect measurement values indicating an illumination intensity in an environment;”; Note: The illumination intensity is one sensed information acquisition related feature.)
It would have been obvious before the effective filing date to combine the AI agents of Mansour and the illumination sensors of Zhang for effective detection of illumination conditions (Zhang, [0087]). Mansour and Zhang are analogous art because they both concern autonomous vehicles.
Claim 3 is rejected over Mansour, Stein, Shen, Lehre and Zhang with the incorporation of claim 1.
Regarding claim 3, Mansour does not teach wherein the at least one sensed information acquisition related feature is related to illumination conditions.
However, Zhang teaches wherein the at least one sensed information acquisition related feature is related to illumination conditions. (Zhang [0063]: “an illumination sensor may collect measurement values indicating an illumination intensity in an environment;”; Note: The illumination intensity is one sensed information acquisition related feature.)
It would have been obvious before the effective filing date to combine the AI agents of Mansour and the illumination sensors of Zhang for effective detection of illumination conditions (Zhang, [0087]). Mansour and Zhang are analogous art because they both concern autonomous vehicles.
Claim 9 is rejected over Mansour, Stein, Shen, Lehre and Zhang with the incorporation of claim 1.
Regarding claim 9, Mansour does not teach autonomously driving the vehicle based on the at least one driving related feature by controlling control [of an acceleration] and speed of the vehicle.
However, Stein teaches autonomously driving the vehicle based on the at least one driving related feature by controlling control of an acceleration [and speed] of the vehicle. (Stein [0181]: “The assessment of each driving response may be computationally processed by assigning numerical scores to each of the criteria, taking into account the current situational assessment. Accordingly, at 2004, speed reduction is assessed according to the various criteria. For example, speed reduction may reduce or prevent splashing, and it may mitigate any harm from striking a potential pothole; however, speed reduction may annoy the vehicle's occupants and, if there is a nearby vehicle that is closely following the autonomous vehicle, speed reduction may cause the nearby vehicle to need to also reduce its speed, increasing the risk of a collision and potentially annoying the driver or occupants of the nearby vehicle.”;)
It would have been obvious before the effective filing date to combine the AI agents of Mansour and the calculation of driving related features of Stein for effective maneuvering of autonomous vehicles (Stein, paragraph [0162]). Mansour and Stein are analogous art because they both concern autonomous vehicles.
Mansour does not teach autonomously driving the vehicle based on the at least one driving related feature by controlling control of an acceleration [and speed] of the vehicle.
However, Zhang teaches autonomously driving the vehicle based on the at least one driving related feature by controlling control of an acceleration [and speed] of the vehicle. (Zhang [0152]: “A reference object may be selected as an object having a high probability in a normal moving state or a stationary state, such as the transportation means 130 on a same traffic lane as target transportation means (the target object) in an environment 100. In a case where the speed difference between a target object and a reference object exceeds a speed difference threshold, if the moving speed of the target object is determined as being higher than that of the reference object, it may be determined that the target object is in an abnormal slow state. Otherwise, if the moving speed of the target object is lower than that of the reference object, it may be determined that the target object is in an abnormal fast state. On some roads, traveling speeds of transportation means need to meet specific speed requirements, and those exceeding a specified upper speed limit or lower speed limit may threaten the safety of other transportation means. In addition, in a moving state, if the speed difference between a transportation means and other transportation means or objects is too large, it is necessary to specially plan and control the driving operation (e.g., deceleration, acceleration, turning, changing the travelling route, etc.) to avoid a collision.”;)
It would have been obvious before the effective filing date to combine the AI agents of Mansour and the speed control of Zhang for effective safe automatic driving (Zhang, [0065]). Mansour and Zhang are analogous art because they both concern autonomous vehicles.
Claim 12 is rejected over Mansour, Stein, Shen, Lehre and Zhang with the incorporation of claim 1.
Regarding claim 12, Mansour teaches wherein at least one other narrow AI agent of the selected sub-set of narrow AI agents is another part of the neural network. (Mansour [col. 5, lines 7-11]: “The HCNN system 10 combines different sub-networks such as the first sub-network 44 and the second sub-network 46 to perform multiple tasks efficiently, thereby using a smaller memory footprint (memory saving) and operating faster than running the different sub-networks separately.”; Note: The second sub-network which is the other narrow AI agent is part of the heterogeneous convolutional neural network (HCNN))
Claim 14 is rejected over Mansour, Stein, Shen, Lehre and Zhang with the incorporation of claim 1.
Regarding claim 14, Mansour does not teach autonomously driving the vehicle based on the at least one driving related feature by controlling control [of an acceleration] and speed of the vehicle.
However, Stein teaches autonomously driving the vehicle based on the at least one driving related feature by controlling control of an acceleration [and speed] of the vehicle. (Stein [0181]: “The assessment of each driving response may be computationally processed by assigning numerical scores to each of the criteria, taking into account the current situational assessment. Accordingly, at 2004, speed reduction is assessed according to the various criteria. For example, speed reduction may reduce or prevent splashing, and it may mitigate any harm from striking a potential pothole; however, speed reduction may annoy the vehicle's occupants and, if there is a nearby vehicle that is closely following the autonomous vehicle, speed reduction may cause the nearby vehicle to need to also reduce its speed, increasing the risk of a collision and potentially annoying the driver or occupants of the nearby vehicle.”;)
It would have been obvious before the effective filing date to combine the AI agents of Mansour and the calculation of driving related features of Stein for effective maneuvering of autonomous vehicles (Stein, paragraph [0162]). Mansour and Stein are analogous art because they both concern autonomous vehicles.
Mansour does not teach autonomously driving the vehicle based on the at least one driving related feature by controlling control of an acceleration [and speed] of the vehicle.
However, Zhang teaches autonomously driving the vehicle based on the at least one driving related feature by controlling control of an acceleration [and speed] of the vehicle. (Zhang [0152]: “A reference object may be selected as an object having a high probability in a normal moving state or a stationary state, such as the transportation means 130 on a same traffic lane as target transportation means (the target object) in an environment 100. In a case where the speed difference between a target object and a reference object exceeds a speed difference threshold, if the moving speed of the target object is determined as being higher than that of the reference object, it may be determined that the target object is in an abnormal slow state. Otherwise, if the moving speed of the target object is lower than that of the reference object, it may be determined that the target object is in an abnormal fast state. On some roads, traveling speeds of transportation means need to meet specific speed requirements, and those exceeding a specified upper speed limit or lower speed limit may threaten the safety of other transportation means. In addition, in a moving state, if the speed difference between a transportation means and other transportation means or objects is too large, it is necessary to specially plan and control the driving operation (e.g., deceleration, acceleration, turning, changing the travelling route, etc.) to avoid a collision.”;)
It would have been obvious before the effective filing date to combine the AI agents of Mansour and the speed control of Zhang for effective safe automatic driving (Zhang, [0065]). Mansour and Zhang are analogous art because they both concern autonomous vehicles.
Claim 4 is rejected under 35 U.S.C. 103 as being unpatentable over Mansour, Stein, Shen and Lehre as applied above and in further view of Baker et al. (US20190050652A1, Obstacle analyzer, vehicle control system, and methods thereof); hereinafter Baker
Claim 4 is rejected over Mansour, Stein, Shen, Lehre and Baker with the incorporation of claim 1.
Regarding claim 4, Mansour does not teach wherein the contextual information includes information pertaining to a driving maneuver to be applied by the vehicle.
However, Baker teaches wherein the contextual information includes information pertaining to a driving maneuver to be applied by the vehicle. (Baker [0057]: “the driving operation may include an autonomous overtaking control (AOC) 260, which may automatically control an overtake maneuver that allows the controlled vehicle to overtake another vehicle (e.g., to overtake a vehicle driving ahead of the controlled vehicle).”; and [0066]: “FIG. 6 shows a schematic flow diagram of a method 600 for controlling a vehicle, according to various aspects. The method 600 may include, in 610, receiving obstacle identification information representing an identification feature of an obstacle in a vicinity of a vehicle (context information); in 620, identifying the obstacle based on the received obstacle identification information and generating an identification value corresponding to the obstacle; in 630, comparing the identification value with one or more reference identification values of a plurality of reference identification values, the plurality of reference identification values representing a plurality of previously identified obstacles, each of the plurality of reference identification values having a rating value assigned thereto; and, in 640, in the case that the identification value matches a reference identification value of the plurality of reference identification values, executing at least one of a triggering or a modification of a driving operation based on the rating value assigned to the reference identification value.”;)
It would have been obvious before the effective filing date to combine the AI agents of Mansour and the overtaking maneuvers of Baker for effective maneuvering of autonomous vehicles when detecting obstacles (Baker [0042]). Mansour and Baker are analogous art because they both concern autonomous vehicles.
Claim 17 is rejected under 35 U.S.C. 103 as being unpatentable over Mansour, Stein, Shen and Lehre in view of Uliyar et al. (US20170337435A1); hereinafter Uliyar in further view of Phadte et al. (US 20200175326 A1); hereinafter Phadte
Claim 17 is rejected over Mansour, Stein, Shen, Lehre, Uliyar and Phadte with the incorporation of claim 1.
Regarding claim 17, Mansour does not teach wherein the different modes of object detection comprise a third mode of object that is slower than a second mode of object detections,
However, Uliyar teaches wherein the different modes of object detection comprise a third mode of object [that is slower than] a second mode of object detections, (Uliyar [0065]: “the ADAS 600 may have a plurality of modes of operation, and the ADAS 600 may operate in one or more of the plurality of modes at a time, based on at least one context. Some non-exhaustive examples of the modes in which the processing system 602 of the ADAS 600 can operate are as follows:
1. Traffic sign detection mode
2. Traffic signal detection mode
3. Road objects detection mode (e.g., Pedestrian detection mode, and Vehicle detection mode)”; Note: 2 is the second mode and 3 is the third mode.)
It would have been obvious before the effective filing date to combine the AI agents of Mansour with the coarse and fine object detection of Uliyar for efficient object detection and scene analysis (Uliyar, [0004]). Mansour and Uliyar are analogous art because they both concern object detection in autonomous vehicles.
Phadte teaches wherein the different modes of object detection comprise a third mode of object that is slower than a second mode of object detections, (Phadte [0004]: While computational requirements are lowered, downsampling these images reduces the range, or distance, of detections due to the fewer number of pixels that are acted upon by the detector. For width and height, for example, the detector may process the image four times as fast, but objects such as cars will be smaller in the downsized image and will need to be twice as close in the camera for them to be the same pixel size, depending on the camera and its field of view (hereinafter “FOV”).”; and [0005]: “As a result, accurate detectors are slower than is typically desirable due to the high computational requirements, while faster detectors using downsampled images are not as accurate as typically desired.”
It would have been obvious before the effective filing date to combine the AI agents of Mansour with the field of view adjustment of Phadte to increase accuracy of object detection (Phadte [0013]). Mansour and Shen are analogous art because they both concern objection detection for autonomous vehicles.
Mansour does not teach has a field of view that is [wider] than a field of view applied during the second mode of object detection, and
However, Phadte teaches has a field of view that is [wider] than a field of view applied during the second mode of object detection, and (Shen [0004]: While computational requirements are lowered, downsampling these images reduces the range, or distance, of detections due to the fewer number of pixels that are acted upon by the detector. For width and height, for example, the detector may process the image four times as fast, but objects such as cars will be smaller in the downsized image and will need to be twice as close in the camera for them to be the same pixel size, depending on the camera and its field of view (hereinafter “FOV”).”;)
It would have been obvious before the effective filing date to combine the AI agents of Mansour with the field of view adjustment of Phadte to increase accuracy of object detection (Shen [0013]). Mansour and Phadte are analogous art because they both concern objection detection for autonomous vehicles.
Uliyar teaches has a field of view that is wider than a field of view applied during the second mode of object detection, and imposes limitations on at least one of distances of interest or objects to be detected. (Uliyar [0025]: “any camera capable of capturing 360 degrees field of view can also be used for the camera system 104.”; and [0026]: “Some non-exhaustive examples of the ‘road objects’ may include other vehicles in front or rear of the vehicle, pedestrians, animals, speed breakers, traffic barriers/barricades, kerbs, lanes, pavements, and the like”)
It would have been obvious before the effective filing date to combine the AI agents of Mansour with the coarse and fine object detection of Uliyar for efficient object detection and scene analysis (Uliyar, [0004]). Mansour and Uliyar are analogous art because they both concern object detection in autonomous vehicles.
Claim 19 is rejected under 35 U.S.C. 103 as being unpatentable over Mansour, Stein, Shen and Lehre in view of Shlens et al. (US 20210279465 A1); hereinafter Shlens
Claim 19 is rejected over Mansour, Stein, Shen, Lehre and Shlens with the incorporation of claim 1.
Regarding claim 19, Mansour does not teach comprising repeating the operating at [different modes] of object detection at different sub-periods of a period that comprises the different sub-periods.
However, Shlens teaches comprising repeating the operating at [different modes] of object detection at different sub-periods of a period that comprises the different sub-periods. (Shlens [0029]: “For each sub-period of the plurality of sub-periods, the system receives current data generated by the sensing system during the sub-period and characterizing a respective partial scene of the environment.”; and [0049]: “In some implementations, the system selects a time length that can facilitate optimal object detection. That is, the system determines the time length for which a measure of accuracy of the current object detection output satisfies a predetermined detection accuracy threshold. FIG. 4 shows example comparisons between sub-periods that are each of different time lengths from each other. In the example of FIG. 4, the predetermined time length is 100 ms and the time lengths for the candidate sub-periods are 25 ms (i.e., if the predetermined time length were to be partitioned into 4 sub-periods), 12.5 ms, 6.25 ms, 3.125 ms, and 1.5625 ms, respectively. In addition, in this example, the measure of accuracy is evaluated using mean average precision (mAP) metric and the predetermined detection accuracy threshold specifies that the object detection outputs must have mAP scores greater than 50.0. Accordingly, in this example, the system can select 25 ms or 12.5 ms as the determined time length for each sub-period.”)
It would have been obvious before the effective filing date to combine the AI agents of Mansour with the sub-periods of Shlens to improve overall efficiency of object detection (Shlens, [0043]). Mansour and Shlens are analogous art because they both concern object detection in autonomous vehicles.
Lehre teaches comprising repeating the operating at different modes of object detection at different sub-periods of a period that comprises the different sub-periods. (Lehre [0004]: “This object is attained by providing a coordinator which, on the one hand, makes available the ranges by the object detection sensor for its respective operating modes in which objects to be detected statistically are to be expected, and on the other hand, informs the coordinator of the driver assistance functions in which sensor recording range parts the objects are to be detected that are required for their functionality. This can be solved by the coordinator in such a way that it correlates the respective distribution density functions with one another, and thus determines the suitable operating mode of the object detection sensor for each driver assistance function, and, corresponding to the instantaneously activated driver assistance function, appropriately switches over the object detection sensor between its operating modes.”)
It would have been obvious before the effective filing date to combine the AI agents of Mansour with the mode switching of Lehre to advantageously operate between multiple modes and accuracies (Lehre [0011]). Mansour and Lehre are analogous art because they both concern object detection in autonomous vehicles.
Conclusion
Applicant's amendment necessitated the new ground(s) of rejection presented in this Office action. Accordingly, THIS ACTION IS MADE FINAL. See MPEP § 706.07(a). Applicant is reminded of the extension of time policy as set forth in 37 CFR 1.136(a).
A shortened statutory period for reply to this final action is set to expire THREE MONTHS from the mailing date of this action. In the event a first reply is filed within TWO MONTHS of the mailing date of this final action and the advisory action is not mailed until after the end of the THREE-MONTH shortened statutory period, then the shortened statutory period will expire on the date the advisory action is mailed, and any nonprovisional extension fee (37 CFR 1.17(a)) pursuant to 37 CFR 1.136(a) will be calculated from the mailing date of the advisory action. In no event, however, will the statutory period for reply expire later than SIX MONTHS from the mailing date of this final action.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to DAVID H TRAN whose telephone number is (703)756-1525. The examiner can normally be reached M-F 9:30 am - 5:30 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, Viker Lamardo can be reached at (571) 270-5871. 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.
/DAVID H TRAN/Examiner, Art Unit 2147
/ERIC NILSSON/Primary Examiner, Art Unit 2151
Prosecution Insights