Notice of Pre-AIA or AIA Status
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
Response to Arguments
Applicant’s arguments and amendment have persuasively overcome the claim objection and most of the 112 rejections.
The remaining issues are addressed below.
112a
Applicant argues:
Applicant submits that an invention need not to solve every problem mentioned in the background to satisfy 112(a).
Examiner responds:
While Applicant’s statement is technically correct, Applicant’s argument misses the issue that the examiner raised. If the specification says that the invention does X, but the invention does not do X, that shows that the inventors had not reduced the invention to practice (either actually or constructively) because, had the inventors reduced the invention to practice, they would know that the invention does not do X.
Applicant argues:
By identifying the rare object, the invention achieves a comprehensive understanding. See Para. [0015].
Examiner responds:
The invention does not identify the rare object, it only performs few shot classification, and thus lacks a comprehensive understanding (i.e., it only has a possible classification). As applicant states for the 101 arguments “The target object class corresponds to objects for which only limited annotated data is available (i.e., "rare objects," see Para. [0047] of the as filed Specification).” Limited annotated data prevents identification, rather the invention classifies (at best).
Applicant argues:
In other words, the target object class can be predetermined, and serve e.g. as a data collection request indicating what objects to be looked for.
Examiner responds:
The present invention is not directed to finding instances of a particular object class (e.g., the car is not directed to drive to find these objects), because recognizing objects from a given target class is distinct from classifying the “nearly endless variety of objects” that the specification is directed to. Specification, [0005].
Applicant argues:
allegedly encompasses SAE Level 5 … including Level 5. These definitions were publicly available and widely recognized by persons skilled in the art at the priority date.
Examiner responds:
The examiner’s understanding is that Applicant is admitting that the claim encompasses level 5 driving.
Applicant argues:
Just because it is not commercially available, does not mean it is not within the skilled person's knowledge.
Examiner responds:
While technically true, that level 5 driving was not commercially available as of the priority date strongly suggests that it was not within ordinary skill. While there may be situations where ordinary skill knowledge is not commercially available, this is not one of them. Rather, many large companies were working hard on level 5, and considered their efforts to be trade secrets (e.g., WAYMO LLC v. Uber Technologies, Inc., 870 F. 3d 1350 (Fed. Cir. 2017)). Thus, level 5 was outside of the skilled person’s knowledge.
Applicant argues:
It is well within the skilled person's understanding that the invention find applicability also in future technologies.
Examiner responds:
Applicant cannot claim technologies that the inventors did not possess.
112b
Applicant argues:
The skilled person readily realizes this as a general term for a system having the functionality to determine and/or monitor a geographical position of the vehicle. Para. [0099] of the specification provides examples of how this can be realized.
Examiner responds:
Claim definiteness is an exacting requirement, and Applicant’s explanation is not precise enough, nor binding.
101
Applicant argues:
Applicant submits that amended independent Claim 1 of the instant application recites steps that are performed by a vehicle and the human mind is not equipped to perform these claimed features. …
Thus, the features of the present claim 1 do not "monopolize every potential solution to the problem,"
Examiner responds:
Applicant is arguing the wrong type of 101 rejection.
Applicant argues:
The present invention addresses a recognized technical problem in automated driving systems: the lack of sufficient training data for rare object classes, where conventional data collection is prohibitively expensive and time-consuming (See para. [0005] of the as filed specification).
Examiner responds:
Under the 112a rejections, Applicant admitted that the present invention does not accomplish the goals of [0005], “Applicant submits that an invention need not to solve every problem mentioned in the background to satisfy 112(a).”
Applicant argues:
This two-stage filtering process allows rapid and cost-efficient identification of relevant data samples
Examiner responds:
This attorney argument is not persuasive. However, if Applicant timely submits evidence (such as a third party sworn statement attesting to the value of this data), that could be persuasive. However, Applicant’s “multi-stage validation process” does not appear to be cost efficient.
Applicant argues:
but are immediately applicable in improving deployed object detection systems in vehicles.
Examiner responds:
At the end of the previous page, Applicant argued that the value of this data is “which can then be incorporated into training datasets.” Incorporating into training dataset is not deploying in vehicles. If Applicant submits timely evidence of deployment in real world vehicles, the examiner will consider it.
Applicant argues:
Thus, the invention implements a multi-stage validation process, where preliminary results are refined and verified before being used in downstream applications such as training dataset formation.
Examiner responds:
The issue is that there are not enough results to be useful. Additionally, the manual verification process undermines Applicant’s arguments that this process is cost effective.
Applicant argues:
to form a training dataset for a machine learning algorithm used in ADS for object identification.
Examiner responds:
The dataset is too small for useful training. See, e.g., “However, a vast majority of these objects such as certain traffic sign types may be exceedingly rare, … .” Specification, [0005].
102
Applicant argues:
Therefore, Lo's runtime classification (Figure 3A) is not based on the "obtained finite support set of annotated images" as explicitly required by Claim 1
Examiner responds:
Applicant is going in an interesting direction, but the present claims are broad enough to encompass Lo’s training. Applicant may wish to narrow the claims to specify that the timing of the obtaining support images is not when the model is being trained.
Applicant argues:
In Lo, the system does not make a "preliminary" association that is later selected for a "subsequent" verification and annotation.
Examiner responds:
Applicant may wish to amend their claims to distinguish parallel and subsequent. Additionally, Lo teaches both a classification and a score.
Claim Objections
Claims 1 and 12 are objected to because of the following informalities:
Claims 1 and 12 recite “each image the one or more images,” but omit “of” after “image” (compare with claim 11).
Appropriate correction is required.
Claim Rejections - 35 USC § 112
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 1-14 (all claims) 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.
All claims are rejected because none of the claims are directed to an invention that provides the benefits articulated at specification [0006]. In particular, the specification teaches few shot learning, see, e.g., [0061], but even if few shot learning is successful, it only provides a general classification, such as determining whether an object is a traffic sign or an animal. Specification, [0040]. This is insufficient for the “comprehensive understanding” that the specification deems “imperative.” Specification, [0005]. Because this specification does not disclose technology that provides the asserted benefits to automated driving (e.g., specification, [0002]), it lacks written description support.
Claims 1, 11 and 12 recite “obtain[ing] a finite support set of annotated images of the target object class,” but this is unlimited functional claiming because the claim has not specified how the system knows which object class is the target class. Because the claim has not specified how the target class is known, the claim covers more than the specification has disclosed. MPEP 2173.05(g). The disclosure at specification, [0065] is insufficient to overcome this rejection because it lacks sufficient detail.
Claim 9 recites “wherein the vehicle comprises an Automated Driving System (ADS).” Specification, [0005] specifically identifies SAE level 5 as being included, but SAE level 5 was not within the level of one of ordinary skill in the art as of the priority date. This is unlimited functional claiming. MPEP 2173.05(g). Applicant may wish to submit evidence to demonstrate claimed level of autonomous driving.
Claims 1, 11, and 12 recite “a threshold number indicating that further samples need to be collected,” but this is new matter. Applicant has not pointed out where the amended claim is supported, nor does there appear to be a written description of this claim limitation in the application as filed. MPEP 2163.04(I)(B). Further, the specification does not even contain a single example of such a threshold (e.g., 3 images are needed for a rare traffic sign from a particular place).
Dependent claims are likewise rejected.
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 1-14 (all claims) are rejected under 35 U.S.C. 112(b) or 35 U.S.C. 112 (pre-AIA ), second paragraph, as being indefinite for failing to particularly point out and distinctly claim the subject matter which the inventor or a joint inventor (or for applications subject to pre-AIA 35 U.S.C. 112, the applicant), regards as the invention.
Claims 1, 11, and 12 recite both “label” and “annotate.” These terms are often synonymous in image recognition, such that the difference between these terms is unclear.
Claims 1, 11, and 12 recite “upon determining that the preliminary association exists,” but it is not clear if this is a timing requirement (i.e., it must happen, and the “upon” specifies when it happens) or if this is a conditional.
Claims 1, 11, and 12 recite “for verifying the preliminary association between the at least one object and the target class,” but it is grammatically unclear which either object or step is modified by “for verifying … .” As an aside, the “for verifying” is an intended use.
Claim 5 recites “the same functionalities,” but this lacks sufficient antecedent basis. MPEP 2173.05(e).
Claim 5 recites “the same functionalities,” but “same” is a relative term and the specification provides insufficient guidance. MPEP 2173.05(b). For example, if one vehicle has less gas than another, does it have the same functionality because both can drive or are they different functionalities because they cannot both drive the same distance (due to the difference in gas).
Claim 5 recites “produced and selected by each vehicle of the plurality of vehicles.” This lacks sufficient antecedent basis because the other vehicles are not recited as having performed the method of claim 1. MPEP 2173.05(e).
Claim 14 recites “localization system,” but this is new terminology. MPEP 2173.05(a). One way to overcome this rejection is to recite specific technology, such as GPS. Specification, [0095].
Dependent claims are likewise rejected.
Claim Rejections - 35 USC § 101
35 U.S.C. 101 reads as follows:
Whoever invents or discovers any new and useful process, machine, manufacture, or composition of matter, or any new and useful improvement thereof, may obtain a patent therefor, subject to the conditions and requirements of this title.
Claims 1-14 (all claims) are rejected under 35 U.S.C. 101 because the claimed invention lacks patentable utility. MPEP 2107.01(I) explains:
Practical utility is a shorthand way of attributing “real-world” value to claimed subject matter. In other words, one skilled in the art can use a claimed discovery in a manner which provides some immediate benefit to the public.
Nelson v. Bowler, 626 F.2d 853, 856, 206 USPQ 881, 883 (CCPA 1980).
However, the utility asserted in the present application does not provide an immediate benefit to the public, but is instead an avenue for further research. MPEP 2107.01(B) discusses In re Fisher, 421 F. 3d 1365 (Fed. Cir. 2005)
The claims at issue in Fisher were directed to expressed sequence tags (ESTs), which are short nucleotide sequences that can be used to discover what genes and downstream proteins are expressed in a cell. The court held that “the claimed ESTs can be used only to gain further information about the underlying genes and the proteins encoded for by those genes. The claimed ESTs themselves are not an end of [the inventor’s] research effort, but only tools to be used along the way in the search for a practical utility…. [Applicant] does not identify the function for the underlying protein-encoding genes. Absent such identification, we hold that the claimed ESTs have not been researched and understood to the point of providing an immediate, well-defined, real world benefit to the public meriting the grant of a patent.” Id. at 1376, 76 USPQ2d at 1233-34). Thus a “substantial utility” defines a “real world” use.
The portions of the present specification that are directed to utility are:
[0002]The disclosed technology relates to methods and systems for determining an association of at least one object to a target object class, wherein the at least one object is present in a surrounding environment of a vehicle travelling on a road. In particular, but not exclusively the disclosed technology relates to recognition and classification of objects being present in the surrounding environment of the vehicle.
[0006]There is thus a pressing need in the art for novel and improved solutions for classification of various objects on roads with high accuracy and speed and without the need for extensive data collection.
[0015]The present inventors have accordingly realized that by using a data-driven approach according to the presented method and systems herein scalability, speed and reproducibility can be achieved in classification of objects such as roadside objects including traffic objects without the stringent requirements on massive data collection. The data-driven approach of the present disclosure provides a flexible, cost-efficient, and rapid approach for generating training data for training neural networks and ML algorithms, specifically for objects for which many samples of real-world data are neither collected nor available. This also greatly contributes to solving the problem of identification of rare objects in scenarios involving multiple environmental variables or conditions happening simultaneously or outside the conventional levels.
[0058]The present inventors have realized that by using a data-driven approach comprising the use of FSL models scalability, speed and reproducibility can be achieved in classification of objects such as roadside traffic objects without the stringent requirements on massive data collection. The data-driven approach of the present disclosure provides a flexible, cost-efficient, and rapid approach for generating training data for training neural networks and ML algorithms, specifically for objects for which many samples of real-world data are neither collected nor available. This also greatly contributes to solving the problem of identification of rare objects in scenarios involving multiple environmental variables or conditions happening simultaneously or outside the conventional levels.
The examiner finds that classification of an object without a precise identification (e.g., knowing that an object is a traffic sign, but not which traffic sign) is not a real world benefit because an autonomous vehicle does not have enough information to act on. The examiner also finds that the alleged advances disclosed in this application are not an immediate benefit, but rather an opportunity for further study, akin to the sequence tags from Fisher. In particular, the specification merely states that this technology generates training data, but does not show what the training data is used for. While the specification identifies a lack of rare objects, the disclosed technology does not address identifying rare objects. Specification:
[0057]The training data is usually obtained through driving the ego vehicle 1, or a plurality of vehicles comprised in a fleet of vehicles, or dedicated test vehicles on various types of roads under a variety of environmental conditions and for suitable periods of time to collect and evaluate large data sets of detected objects on the road. However, the very large variety of objects being present on the road and in the surrounding environment of the vehicle may render the task of data collection and formation of corresponding training data sets practically unattainable. This is even more relevant for scenarios when the object is a rare object such a traffic object e.g. a traffic sign or signal, which has newly been introduced into traffic and for which no adequate amount of data is available yet. Another example may be rare species of animals, which are not completely classified and may drastically vary based on their geographical habitats.
Here, the present technology is not directed to determining the rarity of the object detected, rather it simply identifies the object as a traffic sign or an animal, but this not a benefit for identifying rare objects because there is no guidance produced as to whether there is a rare object, and if so, where it would be categorized.
Here, the present technology allegedly produces classifications of objects on the road, but does not identify why one would want training data that has classified objects as, for example, traffic signs versus animals. Rather, the application has alleged that there is a desire for training date of rare objects (and the examiner finds this persuasive), but fails to teach how rare objects would be distinguished from common objects.
Additionally, even if the above shortcomings were addressed, the Majee reference submitted by Applicant (titled “Few-Shot Learning for Road Object Detection”) closes with “We also observe that class-confusions remains [sic] an open challenge in any few-shot learning paradigm and can be the focus of further improvements.” Comparing the technology discussed in Majee and the present specification does not identify any technology taught in the present specification that overcomes the shortcomings identified in Majee. In other words, even if there were a real world benefit to images objects on or near roads classified with few shot learning, this technology does not resolve the known problem of class confusion.
Therefore, the alleged utility of this invention is not sufficiently specific and substantial.
MPEP 2107.01(I) states:
Practical considerations require the Office to rely on the inventor’s understanding of the invention in determining whether and in what regard an invention is believed to be “useful.” Because of this, Office personnel should focus on and be receptive to assertions made by the applicant that an invention is “useful” for a particular reason.
One way to overcome this rejection may to submit evidence (such as a declaration) showing the real world benefit of the disclosed technology.
Claim Rejections - 35 USC § 102
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 the appropriate paragraphs of 35 U.S.C. 102 that form the basis for the rejections under this section made in this Office action:
A person shall be entitled to a patent unless –
(a)(1) the claimed invention was patented, described in a printed publication, or in public use, on sale, or otherwise available to the public before the effective filing date of the claimed invention.
(a)(2) the claimed invention was described in a patent issued under section 151, or in an application for patent published or deemed published under section 122(b), in which the patent or application, as the case may be, names another inventor and was effectively filed before the effective filing date of the claimed invention.
Claims 1-7 and 9-14 (all claims except for 8) are rejected under 35 U.S.C. 102(a)(1) and/or (a)(2) as anticipated by US20190318207A1 (“Lo”).
1. A method for determining an association of at least one object to a target object class, the at least one object being present in a surrounding environment of a vehicle travelling on a road, (Lo, [0016] “an image classification system of an autonomous or semi-autonomous vehicle”)
the method comprising:
obtaining sensor data from a sensor system of the vehicle, the sensor data comprising one or more images, captured by a vehicle-mounted camera, of the surrounding environment of the vehicle; (Lo, [0046] “For example, the input sensor data 155 can include images captured by a camera of an environment that surrounds a vehicle.” See also Fig. 3C. Fig. 3C shows a stop sign.)
determining a presence of the at least one object in the surrounding environment of the vehicle based on the obtained sensor data; (Lo, [0071] “For instance, the common instance classifier 310 processes the input image 302 to generate a common instance output 304A.”)
obtaining a finite support set of annotated images of the target object class, the finite support set comprising a number of annotated images smaller than a threshold number indicating that further samples need to be collected; (Lo, [0062] “Each of the training examples 123A and 123B includes images of reference objects as well as one or more labels for each image.” Lo’s labels teach the claimed annotations, and the claimed threshold is arbitrarily chosen to be larger than the number of Lo’s images.)
determining, based on a comparison between the obtained sensor data comprising one or more images of the at least one object and the obtained finite support set of annotated images, whether a preliminary association between the at least one object and the target object class exist; (Lo, Fig. 3A, object score 305)
upon determining that the preliminary association exists:
producing an image label for each image the one or more images of the at least one object, thereby obtaining one or more labelled images, the image label indicating the preliminary association of the at least one object with the target object class; and (Lo, Fig. 3A, 305 “object category”)
selecting the one or more labelled images for a subsequent generation of an object annotation for the one or more labelled images for verifying the preliminary association between the at least one object and the target class. (Lo, Fig. 3C. Lo’s score teaches the claimed annotation.)
2. The method according to claim 1, wherein the at least one object comprises at least one road-side traffic object consisting of a traffic sign or a traffic signal. (Lo, Fig. 3C, 302C, stop sign)
3. The method according to claim 1, wherein the method further comprises:
storing the selected one or more labelled images in a memory of the vehicle and/or (Lo, [0102] “Storage devices suitable for tangibly embodying computer program instructions and data include all forms of non-volatile memory … .”)
transmitting the selected one or more labelled images of the at least one object to a remote server and/or (Lo, [0104] “The features can be implemented in a computer system that includes a back-end component, such as a data server … .”)
for generating, at the server, the object annotation based at least on the transmitted one or more labelled images of the at least one object from the vehicle. (Lo, [0102] “Storage devices suitable for tangibly embodying computer program instructions and data include all forms of non-volatile memory … .” The claimed “for a subsequent transmission and “for generating” are both interpreted as intended use. Here, Lo’s system is capable of being used for these intended uses.)
4. The method according to claim 1, wherein the method further comprises: storing the selected one or more labelled images in a memory of the vehicle; and (Lo, Fig. 1, On-Board Classifier Subsystem 134, see also [0049] “… or, in the case of an executing software module, stored within the same memory device.” Lo’s use of “on-board” teaches the claimed memory of the vehicle. Additionally, [0049] is describing the on-board classifier subsystem 134.)
generating the object annotation for the one or more labelled images of the at least one object. (Lo, Fig. 2 and [0069] “classifying the input image in accordance with the determined weight (250)”. Lo’s classifying teaches the claimed annotation.)
5. The method according to claim 4, wherein the method further comprises:
obtaining a set of selected one or more labelled images of the at least one object from a remote server and/or from a plurality of vehicles travelling on the road, comprised in a fleet of vehicles together with the vehicle and having the same functionalities as the vehicle (Lo, [0062] “Each of the training examples 123A and 123B includes images of reference objects as well as one or more labels for each image.” Lo, Fig. 1 shows the training examples 123A and 123B as housed in data center 112.)
wherein the set of selected one or more labelled images of the at least one object comprises the one or more labelled images of the at least one object produced and selected by each vehicle of the plurality of vehicles; (Lo, [0054] “The on-board system 130 can provide the training data 123 to the training system 110 in offline batches or in an online fashion, e.g., continually whenever it is generated.” Lo’s training data teaches the claimed labelled images.)
generating the corresponding object identification annotation based on the obtained set of the selected one or more labelled images of the at least one object. (Lo, Fig. 1)
6. The method according to claim 4, wherein the method further comprises:
forming a training data set for a machine learning, ML, algorithm configured for identification of the at least one object based on the generated object annotation; or (Lo, Fig. 1, training system 110)
transmitting the generated object annotation to a remote server for subsequent forming of the training data set for the machine learning, ML, algorithm, wherein the training data set is formed using the selected one or more labelled images and the generated object annotation. (Lo, Fig. 1, training data 123)
7. The method according to claim 3, wherein the method further comprises:
obtaining, from the remote server, an updated finite support set of annotated images comprising the generated object annotation for the one or more labelled images of the at least one object; or (Lo, Fig. 1, training classifier subsystem 114)
updating the finite support set of annotated images by adding the generated object annotation for the one or more labelled images of the at least one object. (Lo, Fig. 1, training classifier subsystem 114)
9. The method according to claim 1, wherein the vehicle comprises an Automated Driving System (ADS). (Lo, abstract, “an image classification system of an autonomous or semi-autonomous vehicle”)
10. The method according to claim 1, wherein the method is performed by a processing circuitry of the vehicle. (Lo, Fig. 1, vehicle 122 and [0101] “The features described can be implemented in digital electronic circuitry, or in computer hardware, firmware, software, or in combinations of them.”)
Claim 11 is rejected as per claim 1. See also, Lo, claim 15 “One or more non-transitory computer-readable storage media encoded with computer program instructions … .”
Claims 12 and 13 are rejected as per the corresponding method claims. See also, Lo, claim 8, “A system comprising … .”
14. A vehicle comprising: one or more vehicle-mounted sensors configured to monitor a surrounding environment of the vehicle; (Lo, Fig. 1, sensor subsystems 132)
a localization system configured to monitor a geographical position of the vehicle; and (Lo, [0043] “The sensor subsystems include a combination of components that receive reflections of electromagnetic radiation, e.g., LIDAR systems that detect reflections of laser light” Lo’s lidar teaches the claimed monitoring the position of the vehicle, see e.g., [0042] “For example, the vehicle 122 can autonomously apply the brakes if a full-object prediction indicates that a human driver is about to collide with a detected object” because avoiding crashing teaches the claimed monitoring position.)
a system according to claim 12. (See the mapping of claim 12.)
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 text of those sections of Title 35, U.S. Code not included in this action can be found in a prior Office action.
Claim 8 is rejected under 35 U.S.C. 103 as being unpatentable over US20190318207A1 (“Lo”) and US20220189143A1 (“Xie”).
8. Lo teaches the method according to claim 1, but is not relied on for the below claim language.
However, Xie teaches wherein the method further comprises determining whether the preliminary association between the at least one object and the target object class exist by means of a few shot learning model. (Xie, [0004] “Few-shot learning aims at learning to recognize visual categories using only a few labelled exemplars from each category.”)
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to apply the teachings of Xie to the teachings of Lo such that Lo can also classify with few shot learning for the purpose of the advantages from Xie’s background (i.e., Xie, [0003]-[0006]).
Based on the above, this is an example of “combining prior art elements according to known methods to yield predictable results.” MPEP 2143.
Conclusion
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure.
US11953340B2 – titled “Updating road navigation model using non-semantic road feature points”
US12354366B2 – claim 1 “analyze one or more pixels of the at least one image to determine whether the one or more pixels represent at least a portion of a target vehicle, and for pixels determined to represent at least a portion of the target vehicle, determine one or more estimated distance values from the one or more pixels to at least one edge of a face of the target vehicle”
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 ORANGE whose telephone number is (571)270-1799. The examiner can normally be reached Mon-Fri, 9-5.
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, Gregory Morse can be reached at 571-272-3838. 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 ORANGE/ Primary Examiner, Art Unit 2663
Prosecution Insights