Using Machine-Learning Models to Identify and Deter Package Theft

DETAILED ACTION Notice of Pre-AIA or AIA Status 1. 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 Amendments 2. The Amendment filed 07/30/2026 in response to Non-Final Office Action mailed 07/09/2026 has been entered. 3. Claims 1-7, 10-17, and 20-22 are currently pending. Response to Arguments 4. Applicant’s arguments, see pg. 1-5, with regard to the 101 rejections of claims 1-20 have been fully considered and are persuasive. Specifically, the examiner notes that the inclusion of real-time, scene specific machine learning that detects and tracking by training on historical data of a specific property as well as using multi-modal input (image + radar) to perform such detection have obviated the rejection in view of being incapable to perform as a mental process and integrated the judicial exception into a practical application of improvement to camera-based security system technology for package theft deterrence. Therefore, the 101 rejections of claims 1-7, 10-17, and 20 have been withdrawn. 5. Applicant’s arguments, see pg. 6-11, with regard to the 103 rejections of claims 1-7, 10-17, and 20-22 have been fully considered and are partially persuasive. With regard arguments, that Khalili and Madden fail to specifically disclose scene specific model training using historical data from the camera device and at least one other camera device, and identification of both a person and an object along with movement of the object by the person, the examiner disagrees. With regard to identification of both a person and an object along with movement of the object by the person, the examiner notes Khalili explicitly discloses that the system may identify the individual based on shape, movement, and/or defining characteristics, and if said individual is moving the package using the image data ([par. 0067, ln. 1-11]). With regard to historical data training, the examiner notes that Madden uses multiple cameras to determine scene model 123 ([par. 0064, ln. 1-15] see 103 for full citations), and that the historical data used to train the machine learning comprises the scene model ([par. 0069, ln. 1-10], [par. 0070, ln. 1-6], [par. 0071, ln. 1-6] see 103 for full citations), and therefore, Madden discloses scene specific model training using historical data from the camera device and at least one other camera device. Likewise, while Madden teaches in response to movement of the package executing a deterrence action including emitting audiovisual signals in real-time, Madden does not specifically disclose wherein said audiovisual signals are from the camera device ([Fig. 2, see 12:00PM second image 216], [par. 0101, ln. 1-9], [par. 0102, ln. 1-13], [par. 0103, ln. 1-12], [par. 0104, ln. 1-17]). However, the examiner agrees with the applicant with regard to Khalili and Madden failing to teach a local processor, radar sensor, and speaker emitting audiovisual signals from the camera device itself, and likewise, that the identifying of the movement of the object includes radar data from the radar sensor to determine speed, direction or location of movement. Furthermore, arguments with regard to claim 5, 10, 15, and 20 are convincing. Therefore, the 103 rejections have been withdrawn. However, upon further consideration, a new grounds of rejection is made in view of W.O. Publication No. 2022/0245751 to Khalili et al. (hereinafter Khalili), in view of U.S. Publication No. 2023/0282026 to Madden et al. (hereinafter Madden), in view of U.S. Publication No. 2018/0114420 to Siminoff et al. (hereinafter Siminoff), and further in view of U.S. Publication No. 2021/0029542 to Prag et al. (hereinafter Prag). Claim Objections 6. Claims 10 and 20 objected to because of the following informalities: In ln. 3-6, claim 10 recites “…one or more characteristics including… clothing worn… to classify a profession…, and an identification of the sex and approximate age…”. The examiner notes that the “…, and…” language between the limitations makes it unclear if the applicant is intending to recite that both a clothing identification and a sex and approximate age identification are required, which would conflict with the “…one or more…” language of the claim. The examiner enc0ourages amended “…, and…” to “…, or…”. In ln. 3-6, claim 20 recites analogous limitations to claim 10. Appropriate correction is required. Claim Rejections - 35 USC § 103 7. The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action: A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made. The factual inquiries for establishing a background for determining obviousness under 35 U.S.C. 103 are summarized as follows: 1. Determining the scope and contents of the prior art. 2. Ascertaining the differences between the prior art and the claims at issue. 3. Resolving the level of ordinary skill in the pertinent art. 4. Considering objective evidence present in the application indicating obviousness or nonobviousness. 8. Claims 1, 3, 6-7, 11, 13, 16-17 and 22 are rejected under 35 U.S.C. 103 as being unpatentable over W.O. Publication No. 2022/245751 to Khalili et al. (hereinafter Khalili), in view of U.S. Publication No. 2023/0282026 to Madden et al. (hereinafter Madden), in view of U.S. Publication No. 2018/0114420 to Siminoff et al. (hereinafter Siminoff), and further in view of U.S. Publication No. 2021/0029542 to Prag et al. (hereinafter Prag). 9. Regarding Claim 1, Khalili discloses a computer-implemented method comprising ([par. 0007, ln. 1-6] “… a method may identifying one or more package candidates from an image. The method may further include determining an analysis frame from the image based at least in part on the one or more package candidates and executing a shared classifier on the analysis frame, the shared classifier to predict identification of packages and objects within the analysis frame. Further, the method may include outputting a prediction of the identification of the packages and the objects based at least in part on the execution of the shared classifier.”, [par. 0030, ln. 1-6] “The system 102 may include memory 104 and one or more processors 106 coupled to the memory 104. The memory 104 may include one or more computer-readable media. In some embodiments, the computer-readable media may comprise non-transitory computer-readable media. The memory 104 may have one or more instructions stored thereon, wherein the instructions, when executed by the system 102, may cause the system 102 to perform one or more of the operations disclosed herein.”): identifying, by a processor {local to} a camera device comprising an image sensor {and a radar sensor} and executing a machine-learning model, a moveable object located within an area ([par. 0062, ln. 1-13] “In 904, the system may execute a classifier on candidates. For example, the system may execute the classifier on the set of pixels within the candidates (which may be the set of pixels with the removed pixels and the added additional pixels from 210 through 218) to determine whether the area captured by the candidate includes a package. The classifier executed by the system may include a classification model that the system utilizes to predict if the areas captured in each of the candidates includes a package. The classification model may comprise a model produced by machine learning. The classification model may have been trained based on a training set and/or variation set to identify packages within the areas captured in each of the candidates. In some embodiments, the training of the classification model by the system may include supervised learning, where desired results for the elements within the training set and/or the variation set may be indicated to the system… the system may use any machine learning technique (such as random forests, support vector machine, artificial neural networks, or some other machine learning technique) to train the classification model.”), wherein the machine-learning model is a {scene-specific} model trained {by applying the scene-specific model on historical data collected form the camera device and at least one other camera device} ([par. 0062, ln. 1-13]); identifying, by the processor executing the {scene-specific} machine-learning model, a person in the area, an object in the area, and movement of the object by the person using the image data from the image sensor ([par. 0067, ln. 1-11] “…the system may further determine whether there are any other objects that overlap with a predicted package when the system predicts that a package has been delivered. For example, the system may analyze the candidate to determine whether there are any other objects, such as individuals and/or animals within an area of the predicted package. The system may identify the individual and/or animals based on the shape of the objects, the movement of the objects, and/or other defining characteristics of the objects detected within the area of the predicted package… the system may identify an individual within the area of the predicted package, and may determine that the package is being carried by the individual. Based on the determination that the package is being carried by the individual, the system may determine that the package has not been delivered and determine not to initiate a notification in 910 based on the package not having been delivered.”), {wherein identifying the movement of the object by the person includes using radar data from the radar sensor to determine at least one of a speed, a direction, or a location of the movement; and in response to the movement of the object by the person, executing, by the processor executing the scene-specific machine-learning model, a deterrence action, wherein executing the deterrence action includes the processor causing the camera device to emit one or more audiovisual signals via a speaker or a light of the camera device in real time}. Khalili does not specifically disclose a processor local to the camera device including a radar sensor, wherein the machine learning model is scene-specific and trained on historical data from multiple cameras, wherein the movement of the object by the person includes the use of radar data to determine a speed, direction, or location of movement, or wherein a deterrence action is executed by the processor to cause the camera device to emit audiovisual signals from a speaker or a light of the camera device in real-time. However, Madden specifically teaches a scene-specific machine learning model, trained by applying the scene-specific model on historical data collected form the camera device and at least one other camera device ([par. 0064, ln. 1-15] “The scene model database 122 stores a spatial model of one or more scenes where packages may be left for pickup at the property 102. For example, the scene model database 122 can store a model of a scene or scenes including the porch 108. In some examples, the scene model database 122 can build the scene model 123 using images captured by multiple installed cameras with overlapping field of view, e.g., the camera 104 and any other cameras at the property that have a field of view that includes the porch 108. Geometric relationships between the multiple cameras can be stored by the monitoring system 110. The geometric relationships can be determined using image analysis, based on programmed installation positions of the cameras, or both. In some examples, the scene model database 122 can store spatial models of scenes at multiple properties.”, [par. 0069, ln. 1-10] “The package detector 130 can detect that the packages 111, 112 have been left on the porch 108 for pickup based on comparing the first image 115 to the scene model 123… the package detector 130 can perform package detection based at least in part on historical pickup data. For example, a historical pickup database 124 can store data indicating historical package pickup times and locations. The historical package pickup times and locations can be determined, for example, based on historical camera image data and user feedback.”, [par. 0070, ln. 1-6] “…the historical pickup database 124 can store a data entry for each past pickup, with each entry including a time of the pickup and a location of the pickup. The location of the pickup can be, for example, the porch 108, a driveway, a back door, or another location at the property from where a package can be picked up.”, [par. 0071, ln. 1-6] “…the monitoring system 110 may have a training phase. The training phase can be used to generate scene model data for storage in the scene model database 122, historical pickup data for storage in the historical pickup database 124, or both.”, [par. 0072, ln. 1-17] “…the training phase may be a period of multiple days or weeks. During the training phase, the monitoring system 110 can obtain images of all package pickups from the property 102. The monitoring system 110 can store images of pickup personnel and times of the pickups. The monitoring system 110 can also store locations of packages that are left for pickup during the training phase. For example, packages may be left for pickup at the porch 108, at a back door, or at a garage door of the property 102… in addition to or instead of the training phase, the camera 104 and the monitoring system 110 can continuously record and store pickup times, personnel, and locations while in operation. The monitoring system 110 can analyze all pickups, or a selection of pickups… the monitoring system 110 can use electronic pickup notifications received from delivery services as ground truth labeling for the training data.”), and in response to the movement of the object by the person, executing, by the processor executing the scene-specific machine-learning model, a deterrence action, wherein executing the deterrence action includes the processor causing {the camera device} to emit one or more audiovisual signals via a speaker or a light {of the camera device} in real time ([Fig. 2, see 12:00PM second image 216], [par. 0101, ln. 1-9] “FIG. 2 illustrates the example environment 100 in which monitoring system 110 can guard packages from pickup by unauthorized personnel using video. In the example of FIG. 2, a first image 205 shows two packages 211, 212 set out for delivery on the porch 108 at 7:30 am. A second image 216 shows a pickup person 203 approaching the property 102 at 12:00 pm. The pickup person 203 is driving an unmarked sedan 233 and is not wearing a uniform or a nametag.”, [par. 0102, ln. 1-13] “The package monitor 134 performs a verification process and determines that the pickup person 203 is unverified 240… 134 can determine that the pickup person 203 is unverified 240 based on determining that the pickup person 203 does not satisfy any verification criteria… 134 may determine that the sedan 233 does not display any logo or business name that matches an assigned delivery service… 134 may determine that the pickup person 203 is not wearing any uniform or nametag that corresponds to an assigned delivery service… 134 may determine that the pickup person 203 does not provide an accurate password or code.”, [par. 0103, ln. 1-12] “… 134 can determine that the pickup person 203 is unverified 240 based on determining that the pickup person 203 satisfies fewer than a required number of verification criteria. For example, the required number of verification criteria may be two verification criteria. The pickup person 203 may satisfy a first verification criterion by arriving within the expected pickup time window, but might not satisfy any second criterion. Thus, due to the pickup person 203 not satisfying at least the required number of verification criteria… 134 can determine that the pickup person 203 is unverified.”, [par. 0104, ln. 1-17] “When the pickup person 203 approaches the property 102… 134 can provide a warning 236 based on determining that the pickup person 203 is unverified 240. The warning 236 can include, for example, a verbal or visual prompt, e.g., challenge, presented through the signaling device 242. The prompt can include a request to provide a credential, e.g., by speaking the person's name, by speaking a password, or by presenting a visible code to the camera 104. If the pickup person 203 does not provide the requested credentials… 134 can provide an additional warning 236, e.g., including a verbal or textual instruction. The instruction may state, for example, “put the package down.” If the pickup person 203 picks up one of the packages without being verified… 134 can escalate the warning 236 to an alarm. The alarm can include, for example, a siren sound and/or a flashing light.”). One of ordinary skill in the art, before the effective filing date of the claimed invention, would recognize Khalili and Madden as within the same field of video processing for detecting and managing delivery packages, and as analogous to the claimed invention. With regard to the scene-specific machine learning model and training, the motivation to combine would have been obvious to one of ordinary skill in the art, and is taught in Madden, wherein the model may be improved over time via continual usage of historical data ([par. 0066, ln. 1-6] “…the scene model database 122 can generate, build, or improve the scene model 123 over time. The scene model database 122 can build the model of the scene using images captured by the camera 104 and by any other cameras installed at the property 102.”). With regard to the deterrence action, the motivation to combine is disclosed in Madden, wherein the deterrence action prevents unauthorizes people from stealing packages ([par. 0101, ln. 1-9]). One of ordinary skill in the art, before the effective filing date of the claimed invention, would have combined the method of Khalili with the scene-specific machine learning model, historical data training, and package movement deterrence action including audiovisual signals of Madden through known means, with no change to their respective function, and the combination would had yielded nothing more than predicable results. Specifically, one of ordinary skill in the art would have combined the method of Khalili with the scene-specific machine learning model, historical data training, and package movement deterrence action including audiovisual signals of Madden such that the machine learning model of Khalili is trained as a scene-specific model using historical data and the processor generates a deterrence action including an audiovisual signal if an unauthorized individual is detected moving the package. Madden does not specifically disclose a processor local to the camera device including a radar sensor, wherein the movement of the object by the person includes the use of radar data to determine a speed, direction, or location of movement, or wherein the camera device emits audiovisual signals from a speaker or a light of the camera device. Specifically, the examiner notes that while Madden teaches that audiovisual signals including a speaker and a light may be emitted, Madden discloses these as separate device connected to a processor as opposed to components within the camera device itself. However, Siminoff specifically discloses a processor local to the camera device, and wherein the camera device contains both a speaker and a light for emitting audiovisual signals including deterrence actions ([Fig. 3, see Speaker 151, Camera 154 and image processor 240, controller processor 160, Light indicators 162], [par. 0132, ln. 1-17] “The LEDs 162 and the light pipe 232 may function as visual indicators for a visitor and/or a user. For example, the LEDs 162 may illuminate upon activation or stay illuminated continuously. In one aspect, the LEDs 162 may change color to indicate that the front button 148 has been pressed. The LEDs 162 may also indicate that the battery 142 needs recharging, or that the battery 142 is currently being charged, or that charging of the battery 142 has been completed. The LEDs 162 may indicate that a connection to the user's wireless network is good, limited, poor, or not connected. The LEDs 162 may be used to guide the user through setup or installation steps using visual cues, potentially coupled with audio cues emitted from the speaker 152.” [par. 0194, ln. 1-23] “the alert may comprise an audible alarm emitted from the speaker 152 of the wireless A/V recording and communication device 130. The audible alarm may be any loud noise likely to attract attention and/or startle the unauthorized person, making it more likely that he or she will flee without absconding with the parcel(s). In some embodiments, the alert may comprise an announcement emitted from the speaker 152 of the wireless A/V recording and communication device 130. The announcement may comprise a verbal warning that the area about the wireless A/V recording and communication device 130 is being recorded. The unauthorized person, upon being informed that the area about the wireless A/V recording and communication device 130 is being recorded, may decide to flee the scene without absconding with the parcel(s). In some embodiments, the alert may comprise both an audible alarm and an announcement in combination. Also in some embodiments, the alert may comprise any combination of an alert signal sent to a client device, an audible alarm emitted from the speaker 152 of the wireless A/V recording and communication device 130, and an announcement emitted from the speaker 152 of the wireless A/V recording and communication device 130.”). One of ordinary skill in the art, before the effective filing date of the claimed invention, would recognize Khalili, Madden, and Siminoff as within the same field of video processing for detecting and managing delivery packages, and as analogous to the claimed invention. The motivation to combine would have been obvious to one of ordinary skill in the art, in that by incorporating the processor, speaker, and light in a local manner with the camera device/image sensor, delays and issues resulting from the need to communicate between various components can be minimized (e.g., if the system is a distributed system analogous to that disclosed in Madden, connection may be severed between processor and video in case of internet outage or the like, which can be prevented by connecting the components physically within a single device analogous to Siminoff). One of ordinary skill in the art, before the effective filing date of the claimed invention, would have combined the method of Khalili with the scene-specific machine learning model, historical data training, and package movement deterrence action including audiovisual signals of Madden, and further combined the method of the combination of Khalili and Madden with the local processor, speaker, and light as taught in Siminoff, through known means, with no change to their respective function, and the combination would had yielded nothing more than predicable results. Specifically, one of ordinary skill in the art would have combined the method of the combination of Khalili and Madden with Siminoff such that the method of Khalili and Madden were performed on a singular device including a local processor, image sensor, speaker, and light analogous so Siminoff to prevent delays and network issues that may occur from distributed systems. However, Siminoff does not specifically disclose a radar sensor, or wherein the movement of the object by the person includes the use of radar data to determine a speed, direction, or location of movement. However, Prag specifically teaches wherein a camera device includes a local radar sensor ([Fig. 2, see integrated Radar circuit 202], [par. 0060, ln. 1-6] “This example of the authentication system 114 is illustrated showing an interior 200 of the UE 102 (shown as a smartphone). In the depicted configuration, the UE 102 includes a radar integrated circuit 202 of the radar system 104, a speaker 204, a front-facing camera 206, a proximity sensor 208, and an ambient light sensor 210.”), and wherein tracking movement of a target object includes the use of radar data to determine a speed, direction, or location of movement ([par. 0038, ln. 1-20] “The reflection from the user 120 in the radar field 118 enables the radar system 104 to determine various information about the user 120, such as the body position and posture of the user 120, which may indicate a variety of different nonverbal body language cues, body positions, or body postures. The cues, positions, and postures may include an absolute position or distance of the user 120 with reference to the UE 102, a change in the position or distance of the user 120 with reference to the UE 102 (e.g., whether the user 120 or the user's hand or object held by the user 120 is moving closer to or farther from the UE 102), the velocity of the user 120 (e.g., a hand or a non-user object) when moving toward or away from the UE 102, whether the user 120 turns toward or away from the UE 102, whether the user 120 leans toward, waves toward, reaches for, or points at the UE 102, and so forth. These reflections can also be analyzed to determine, or to add confidence to, authentication, such as an identity of a human through analysis of the radar data (e.g., scattering centers of a user's face).”, [par. 0039, ln. 1-21] “The radar manager 106 is configured to determine, based on radar data from the radar system 104, a user's intent to engage, disengage, or maintain engagement with the UE 102. A user's intent can be deduced from the various cues, positions, postures, and distances/velocities noted above, such as based on a hand or arm reach toward, a movement of eyes to look at, or movement of a head or face oriented toward the UE 102. For a hand or arm reach, the radar manager 106 determines that the user is reaching their hand or orienting their arm in such a way as to indicate a likely intent to touch or pick up the UE 102. Examples include a user reaching toward a volume button on a wirelessly attached speaker, a reach toward a wireless or wired mouse associated with a tablet computer, or a reach toward the UE 102 itself This reach toward can be determined based on a hand movement alone, an arm and hand movement, or an arm bending or straightening in a manner that permits a hand of the arm to touch or grab the UE 102. As noted in FIGS. 14-16 below, this intent to engage determination can be for non-users or users, authenticated or not.”). One of ordinary skill in the art, before the effective filing date of the claimed invention, would recognize Khalili, Madden, Siminoff, and Prag as within the same field of video processing for object tracking and action detection, and as analogous to the claimed invention. The motivation to combine is specifically disclosed in Prag, wherein it allows for reduced power consumption ([par. 0049, ln. 9-16] “…while the radar manager 106 may determine a user's intent and the movement manager 110 can determine movement, some of which are determined to indicate a user's intent to engage with the UE 102, the state manager 112, by using both of their determinations, can improve the accuracy, robustness, and/or speed of an overall determination that the user's intent is to engage, disengage, or maintain engagement with the UE 102.”, [par. 0051, ln. 1-6] “In some cases, however, the state manager 112 determines to increase power or otherwise prepare the state of the UE 102 responsive to both inertial data and radar data, e.g., the radar manager 106 determining that the user is intending to engage and the movement manager 110 determining that the user is picking up the UE 102.”). One of ordinary skill in the art, before the effective filling date of the claimed invention, would specifically recognize this is particularly advantageous to package monitoring, wherein long periods of time may occur between delivery and movement of the package by a user or unauthorized individual (e.g., by using radar instead of video to detect movement, the image sensor of the camera device may only be activated when radar detects movement within a region of a package). One of ordinary skill in the art, before the effective filing date of the claimed invention, would have combined the method of the combination of Khalili, Madden, and Siminoff with the integrated radar sensor and radar speed, direction, and location of movement detection of Prag through known means, with no change to their respective function, and the combination would had yielded nothing more than predicable results. Specifically, one of ordinary skill in the art would have combined the method of the combination of Khalili, Madden, and Siminoff such that it was performed including an integrated radar sensor and radar speed, direction, and location of movement detection of Prag to determine movement of the object by the person. Therefore, it would have been obvious to one of ordinary skill in the art, before the effective filing date of the claimed invention, to combine the method of Khalili with the scene-specific machine learning model, historical data training, and package movement deterrence action including audiovisual signals of Madden, the local processor, speaker, and light of Siminoff, and the integrated radar sensor and radar speed, direction, and location of movement detection of Prag to obtain the invention as specified in claim 1. 10. Regarding Claim 3, a combination of Khalili, Madden, Siminoff, and Prag teaches the method of claim 1. Khalili further discloses wherein identifying the moveable object includes determining an object category of the object ([par. 0067, ln. 1-11], [par. 0068, ln. 1-7] “…the classifier may include a classification model that the system utilizes to predict if an area within the predicted packages includes an individual or an animal. The classification model may be implemented with the same classification model as the classification model used for predicting the packages or may be a separate classification model from the classification model used for predicting the packages. The classification model may have been trained based on a training set and/or variation set to identify individuals and/or animals within the areas captured in each of the candidates.”). Specifically, one of ordinary skill in the art, before the effective filing date of the claimed invention, would recognize Khalili discloses wherein the movable objects can further be categorized as an animal or person. Therefore, it would have been obvious to one of ordinary skill in the art, before the effective filing date of the claimed invention, to combine the method of Khalili with the scene-specific machine learning model, historical data training, and package movement deterrence action including audiovisual signals of Madden, the local processor, speaker, and light of Siminoff, and the integrated radar sensor and radar speed, direction, and location of movement detection of Prag to obtain the invention as specified in claim 3. 11. Regarding Claim 6, a combination of Khalili, Madden, Siminoff, and Prag teaches the method of claim 1. Khalili and Madden do not specifically teach wherein identifying the movement of the object by the person includes determining a direction of the movement. However, Siminoff specifically teaches wherein identifying the movement of the object by the person includes determining a direction of the movement using image data ([par. 0180, ln. 1-24] “…determining whether removal of the parcel from the area about the wireless A/V recording and communication device 130 was authorized may comprise detecting (or tracking) a direction of movement of the parcel. For example, when a parcel is left outside the front entrance of a home, the homeowner (or other occupant) will typically pick up the parcel and bring it inside the home. A parcel thief, by contrast, will typically pick up the parcel and carry it away from the home. Thus, if the wireless A/V recording and communication device 130 detects that the parcel is moving toward a structure to which the wireless A/V recording and communication device 130 is secured (or with which the wireless A/V recording and communication device 130 is associated), then the process may determine that the removal of the parcel from the area about the wireless A/V recording and communication device 130 is authorized. But, if the wireless A/V recording and communication device 130 detects that the parcel is moving away from the structure to which the wireless A/V recording and communication device 130 is secured (or with which the wireless A/V recording and communication device 130 is associated), then the process may determine that the removal of the parcel from the area about the wireless A/V recording and communication device 130 is unauthorized.”). Likewise, Prag teaches direction of movement determination, though the examiner notes this is performed using radar data as opposed to image data ([par. 0038, ln. 1-20], [par. 0039, ln. 1-21]). The motivation to combine is disclosed in Siminoff ([par. 0180, ln. 1-24]), wherein tracking the direction of the movement is helpful in determining if a movement of the object is a suspicious activity (e.g., package theft). One of ordinary skill in the art, before the effective filling date of the claimed invention, would have combined the method of the combination of Khalili and Madden with the local processor, speaker, light, and direction of movement detection of Siminoff, and the radar and radar speed, direction, and location of movement detection of Prag through known means, with no change to their respective function, and the combination would have yielded nothing more than predicable results. Specifically, one of ordinary skill in the art, in combining Khalili, Madden, Siminoff, and Prag, would have incorporated the image based direction of movement as taught in Siminoff to prevent package theft. Therefore, it would have been obvious to one of ordinary skill in the art, before the effective filing date of the claimed invention, to combine the method of Khalili with the scene-specific machine learning model, historical data training, and package movement deterrence action including audiovisual signals of Madden, the local processor, speaker, and light of Siminoff, and the integrated radar sensor and radar speed, direction, and location of movement detection of Prag to obtain the invention as specified in claim 6. 12. Regarding Claim 7, a combination of Khalili, Madden, Siminoff, and Prag teaches the method of claim 6. Rejections analogous to claim 6 are further applicable to claim 7. Specifically, Khalili does not specifically disclose executing the deterrence action includes determining that the direction of the movement is away from a building associated with the area. Madden teaches executing the deterrence action ([Fig. 2], [par. 0101, ln. 1-9], [par. 0102, ln. 1-13], [par. 0103, ln. 1-12], [par. 0104, ln. 1-17]), but does not specifically disclose that this is in response to determining that the direction of the movement is away from a building associated with the area, only in response to an unauthorized person picking up the package. Prag does not specifically disclose a deterrence action. However, Siminoff specifically teaches executing a deterrence action in response to determining that the direction of the movement is away from a building associated with the area ([par. 0180, ln. 1-24], [par. 0194, ln. 1-23] “…the alert may comprise an audible alarm emitted from the speaker 152 of the wireless A/V recording and communication device 130. The audible alarm may be any loud noise likely to attract attention and/or startle the unauthorized person, making it more likely that he or she will flee without absconding with the parcel(s)… the alert may comprise an announcement emitted from the speaker 152 of the wireless A/V recording and communication device 130. The announcement may comprise a verbal warning that the area about the wireless A/V recording and communication device 130 is being recorded. The unauthorized person, upon being informed that the area about the wireless A/V recording and communication device 130 is being recorded, may decide to flee the scene without absconding with the parcel(s)… the alert may comprise both an audible alarm and an announcement in combination… the alert may comprise any combination of an alert signal sent to a client device, an audible alarm emitted from the speaker 152 of the wireless A/V recording and communication device 130, and an announcement emitted from the speaker 152 of the wireless A/V recording and communication device 130.”). The motivation to combine remains analogous to claim 6. One of ordinary skill in the art, before the effective filling date of the claimed invention, would have combined the method of the combination of Khalili and Madden with the local processor, speaker, light, and direction of movement detection of Siminoff, and the radar and radar speed, direction, and location of movement detection of Prag through known means, with no change to their respective function, and the combination would have yielded nothing more than predicable results. Specifically, one of ordinary skill in the art, in combining Khalili, Madden, Siminoff, and Prag, would have further identified the movement of the object by an unauthorized person away from the house as a suspicious activity, and executed a deterrence action in response as taught in Siminoff. Therefore, it would have been obvious to one of ordinary skill in the art, before the effective filing date of the claimed invention, to combine the method of Khalili with the scene-specific machine learning model, historical data training, and package movement deterrence action including audiovisual signals of Madden, the local processor, speaker, and light of Siminoff, and the integrated radar sensor and radar speed, direction, and location of movement detection of Prag to obtain the invention as specified in claim 7. 13. Regarding Claim 22, a combination of Khalili, Madden, Siminoff, and Prag teaches the method of claim 1. Khalili and Madden do not specifically teach wherein identifying the movement of the object by determining at least one of a speed, a direction, or a location of the movement further comprises the local processor of the camera device determining a velocity and direction of the object relative to a building associated with the area being monitored to confirm that a deterrence action is to be executed, where the movement and direction is determined to be away from the building, or otherwise determining that the object is being stowed or put away in the building an no deterrence action is warranted, where the movement and direction is determined to be toward the building. However, Siminoff specifically teaches wherein identifying the movement of the object by determining at least a direction of the movement further comprises the local processor of the camera device determining {a velocity} and direction of the object relative to a building associated with the area being monitored to confirm that a deterrence action is to be executed, where the movement and direction is determined to be away from the building ([par. 0180, ln. 1-24] “…For example, when a parcel is left outside the front entrance of a home, the homeowner (or other occupant) will typically pick up the parcel and bring it inside the home. A parcel thief, by contrast, will typically pick up the parcel and carry it away from the home. Thus, if the wireless A/V recording and communication device 130 detects that the parcel is moving toward a structure to which the wireless A/V recording and communication device 130 is secured (or with which the wireless A/V recording and communication device 130 is associated), then the process may determine that the removal of the parcel from the area about the wireless A/V recording and communication device 130 is authorized…”, [par. 0194, ln. 1-23]), or otherwise determining that the object is being stowed or put away in the building an no deterrence action is warranted, where the movement and direction is determined to be toward the building ([par. 0180, ln. 1-24], [par. 0194, ln. 1-23]). Specifically, Siminoff performs this direction of movement detection using images as opposed to radar, and does not specifically disclose a velocity is determined. The motivation to combine remains analogous to claim 6. One of ordinary skill in the art, before the effective filling date of the claimed invention, would have combined the method of the combination of Khalili and Madden with the local processor, speaker, light, and direction of movement detection of Siminoff through known means, with no change to their respective function, and the combination would have yielded nothing more than predicable results. Specifically, one of ordinary skill in the art, before the effective filling date of the claimed invention, would have combined Khalili, Madden, and Siminoff such that a direction of movement was away or toward the building is used to trigger a deterrence action as taught in Siminoff. However, Siminoff does not specifically disclose a velocity determination. Therefore, a combination of Khalili, Madden, and Siminoff does not specifically disclose a determining a velocity of the movement of the object. However, Prag specifically discloses wherein analogous speed, direction, and movement of an object can be performed by radar, and wherein the velocity of an object is determined using radar ([Fig. 2, see integrated Radar circuit 202], [par. 0060, ln. 1-6], [par. 0038, ln. 1-20], [par. 0039, ln. 1-21]). The motivation to combine remains analogous to claim 1, wherein integrated radar allows for reduced power consumption ([par. 0049, ln. 9-16], [par. 0051, ln. 1-6]). One of ordinary skill in the art, before the effective filling date of the claimed invention, would have combined the method of the combination of Khalili, Madden, and Siminoff with the radar based speed, direction, and movement determination and determining a velocity of the movement of the object of Prag through known means, with no change to their respective function, and the combination would have yielded nothing more than predicable results. Specifically, one of ordinary skill in the art, in combining the method of the combination of Khalili, Madden, and Siminoff with Prag would have further determined at least one of a speed, a direction, or a location of the movement and a velocity of an object using the radar sensor. Therefore, it would have been obvious to one of ordinary skill in the art, before the effective filing date of the claimed invention, to combine the method of Khalili with the scene-specific machine learning model, historical data training, and package movement deterrence action including audiovisual signals of Madden, the local processor, speaker, and light of Siminoff, and the integrated radar sensor and radar speed, direction, and location of movement detection of Prag to obtain the invention as specified in claim 7. 14. Regarding Claim 11, the claim language is analogous to claim 1, directed toward the camera device with a local processor, radar sensor, speaker, and light as described in the method of claim 1. Rejections analogous to claim 1 are further applicable to claim 11 in view of the device described in claim 1. Specifically, the examiner notes Khalili likewise discloses a camera device ([Fig. 1, see 112]), and that with regard to the remainder of the claim, rejections analogous to claim 1 are further applicable. Therefore, it would have been obvious to one of ordinary skill in the art, before the effective filing date of the claimed invention, to combine the camera device of Khalili with the scene-specific machine learning model, historical data training, and package movement deterrence action including audiovisual signals of Madden, the local processor, speaker, and light of Siminoff, and the integrated radar sensor and radar speed, direction, and location of movement detection of Prag to obtain the invention as specified in claim 11. 15. Regarding Claims 13, and 16-17, a combination of Khalili, Madden, and Siminoff teaches the camera device of claim 11. The claim language of claims 13, and 16-17 is analogous to claims 3, and 6-7 respectively. Rejections analogous to claims 3, and 6-7 are further applicable to claims 13, and 16-17 in view of the camera device of the combination of Khalili, Madden, Siminoff, and Prag. Therefore, it would have been obvious to one of ordinary skill in the art, before the effective filing date of the claimed invention, to combine the camera device of Khalili with the scene-specific machine learning model, historical data training, and package movement deterrence action including audiovisual signals of Madden, the local processor, speaker, and light of Siminoff, and the integrated radar sensor and radar speed, direction, and location of movement detection of Prag to obtain the invention as specified in claims 13 and 16-17. 16. Claim 2, 4, 12, and 14 are rejected under 35 U.S.C. 103 as being unpatentable over W.O. Publication No. 2022/245751 to Khalili, in view of U.S. Publication No. 2023/0282026 to Madden, in view of U.S. Publication No. 2018/0114420 to Siminoff, and further in view of U.S. Publication No. 2021/0029542 to Prag, and further in view of U.S. Publication No. 2018/0349708 to van Hoof et al. (hereinafter van Hoof). 17. Regarding Claim 2, a combination of Khalili, Madden, Siminoff, and Prag teaches the method of claim 1. Khalili, Madden, Siminoff, and Prag do not specifically teach wherein identifying the moveable object includes tracking movement of the object into the area. However, van Hoof specifically teaches wherein identifying the movable object includes tracking movement of the object into the area ([par. 0191, ln. 1-14] “…motion has been detected more than a threshold number of times in the first identified region of interest. A threshold number of times that motion is detected at an area of the field of view (more particularly, going into or out of the area) may be predefined, learned (e.g., from video/event history), etc. at the camera 118 or video server system 508. The camera 118 or video server system 508 may determine that a certain area of the field of view satisfies the threshold and determine that the area is a source (number of times motion detected going into the area satisfies the threshold) or sink (number of times motion detected going out of the area satisfies the threshold) of motion, and identify the source/sink area as a region of interest.”, [par. 0210, ln. 1-7] “…the detected motion includes motion starting from a region of interest and/or motion starting from an ingress/egress area. Detected motion may originate from a source area and/or terminate in a sink area, and the source/sink area may be identified as a region of interest based on motion originating or terminating in the area.”, [par. 0211, ln. 1-6] “…a region of interest may be identified based on automatic detection of objects in the field of view (e.g., region of interest created on mail package, region of interest created on car entering driveway); the region of interest is automatically identified based on the automatic object detection.”, [par. 0212, ln. 1-6] “…a region of interest is identified for a detected person or face associated with an automatically created region of interest (e.g., a stranger stealing a package taken out of an automatically created region of interest, a person entering or exiting a region of interest).”). One of ordinary skill in the art, before the effective filing date of the claimed invention, would recognize Khalili, Madden, Siminoff, Prag, and van Hoof as within the same field of video processing for object tracking and action detection, and Khalili, Madden, Siminoff, and van Hoof as further within the same field of managing delivery packages, and as analogous to the claimed invention. The motivation to combine would have been obvious to one of ordinary skill in the art, and is disclosed in van Hoof ([par. 0212, ln. 1-6]), in that by tracking movement of the object into and out of the area, you can ascertain theft of a package. One of ordinary skill in the art, before the effective filing date of the claimed invention, would have combined the method of the combination of Khalili, Madden, Siminoff, and Prag with the movement tracking into an area of van Hoof, through known means, with no change to their respective function, and the combination would had yielded nothing more than predicable results. Specifically, one of ordinary skill in the art would have modified the method of the combination of Khalili, Madden, Siminoff, and Prag to implement the movement tracking into an area of van Hoof to identify the object by tracking movement of the object (e.g., a package) into the area (e.g., from a source, like driveway, walkway, etc.). Therefore, it would have been obvious to one of ordinary skill in the art, before the effective filing date of the claimed invention, to combine the method of Khalili with the scene-specific machine learning model, historical data training, and package movement deterrence action including audiovisual signals of Madden, the local processor, speaker, and light of Siminoff, the integrated radar sensor and radar speed, direction, and location of movement detection of Prag, and the movement tracking into an area of van Hoof to obtain the invention as specified in claim 2. 18. Regarding Claim 4, a combination of Khalili, Madden, Siminoff, and Prag teaches the method of claim 1. Khalili, Madden, Siminoff, and Prag do not specifically disclose wherein identifying the movement of the object by the person includes determining that the person has entered the area. However, van Hoof specifically teaches identifying the movement of the object by the person includes determining that the person has entered the area ([par. 0191, ln. 1-14], [par. 0210, ln. 1-7], [par. 0211, ln. 1-6], [par. 0212, ln. 1-6]). Specifically, the motivation to combine remains analogous to claim 2, and is disclosed in van Hoof ([par. 0212, ln. 1-6]), wherein tracking a person entering and leaving an area assists in determination of package theft. One of ordinary skill in the art, before the effective filing date of the claimed invention, would have combined the method of the combination of Khalili, Madden, Siminoff, and Prag with the movement tracking into an area of van Hoof, through known means, with no change to their respective function, and the combination would had yielded nothing more than predicable results. Specifically, one of ordinary skill in the art would have modified the method of the combination of Khalili, Madden, Siminoff, and Prag to implement the determination that a person has entered the area of van Hoof to identify movement of the object by a person by tracking movement of the person into the area (e.g., from a source to the package, to a sink with the package, etc.). Therefore, it would have been obvious to one of ordinary skill in the art, before the effective filing date of the claimed invention, to combine the method of Khalili with the scene-specific machine learning model, historical data training, and package movement deterrence action including audiovisual signals of Madden, the local processor, speaker, and light of Siminoff, the integrated radar sensor and radar speed, direction, and location of movement detection of Prag, and the movement tracking into an area of van Hoof to obtain the invention as specified in claim 4. 19. Regarding Claims 12 and 14, a combination of Khalili, Madden, Siminoff, and Prag teaches the camera device of claim 11. The claim language of claims 12 and 14 is analogous to claims 2 and 4 respectively. Rejections analogous to claims 2 and 4 are further applicable to claims 12 and 14 in view of the camera device of the combination of Khalili, Madden, Siminoff, and Prag. Therefore, it would have been obvious to one of ordinary skill in the art, before the effective filing date of the claimed invention, to combine the camera device of Khalili with the scene-specific machine learning model, historical data training, and package movement deterrence action including audiovisual signals of Madden, the local processor, speaker, and light of Siminoff, the integrated radar sensor and radar speed, direction, and location of movement detection of Prag, and the movement tracking into an area of van Hoof to obtain the invention as specified in claims 12 and 14. 20. Claim 21 is rejected under 35 U.S.C. 103 as being unpatentable over W.O. Publication No. 2022/245751 to Khalili, in view of U.S. Publication No. 2023/0282026 to Madden, in view of U.S. Publication No. 2018/0114420 to Siminoff, and further in view of U.S. Publication No. 2021/0029542 to Prag, and further in view of U.S. Publication No. 2023/0410519 to Koga (hereinafter Koga). 21. Regarding Claim 21, a combination of Khalili, Madden, Siminoff, and Prag teaches the method of claim 1. Khalili, Madden, and Prag do not specifically disclose executing the deterrence action includes executing the deterrence action based on one or more determined actions of the person, including: the person attempting to hide from the camera device; the person passing by a building associated with the area being monitored multiple times; and the person looking in the windows of the building. Siminoff teaches wherein executing the deterrence action includes executing the deterrence action based on one or more determined actions of the person ([par. 0180, ln. 1-24], [par. 0194, ln. 1-23]). Specifically, Siminoff teaches wherein the deterrence action can be triggered by a person picking up the object and walking away from the building, as well as “suspicious events”, but does not specifically teach wherein the determined actions include the person attempting to hide from the camera device; the person passing by a building associated with the area being monitored multiple times; and the person looking in the windows of the building. However, Koga specifically teaches wherein a deterrence action is triggered based on one or more determined actions of a person, including the person passing by an area being monitored multiple times ([par. 0081, ln. 1-15] “…Each of the cameras C1, . . . can determine whether an alarm notification relating to the person is to be issued by calculating the crime index (score) indicating the degree of the sign of the crime based on the analysis result (specifically, the emotion index and the behavior analysis result of the person).”, [par. 0084, ln. 1-12] “The Table TB1 illustrates a list of the suspicious behavior examples for determining whether a person corresponding to the second time series data performs a suspicious behavior based on the analysis result of the second time series data analyzed by using the fourth trained AI model. Although not illustrated in the table TB1 illustrated in FIG. 6, a specific condition (a threshold value) such as “number of times of passage” and “time of passing through the same place” may be set as an additional condition in a condition for determining a behavior to be a suspicious behavior, for example, “pass through the same place several times over a long time””, [par. 0085, ln. 1-10] “The camera C1 analyzes the second time series data by using the fourth trained AI model, and increases a value of the crime index (score) in response to determining that the person corresponding to the second time series data performs a suspicious behavior such as “repeatedly view direction of salesperson and the other customers”, “pass through the same place several times over a long time”, “confirm monitoring camera several times”, “confirm rear several times”, or “a plurality of persons are gathered” illustrated in the table TB1.”). One of ordinary skill in the art, before the effective filing date of the claimed invention, would recognize Khalili, Madden, Siminoff, Prag, and Koga as within the same field of video processing for object tracking and action detection, and as analogous to the claimed invention. The motivation to combine is disclosed in Koga, wherein such actions can be used to determine if a crime is being committed or is soon to be committed ([par. 0081, ln. 1-15]). One of ordinary skill in the art, before the effective filing date of the claimed invention, would have combined the method of the combination of Khalili, Madden, Siminoff, and Prag with the determined action detection of Koga through known means, with no change to their respective function, and the combination would have yielded nothing more than predicable results. Specifically, one of ordinary skill in the art would have combined the method of the combination of Khalili, Madden, Siminoff, and Prag with the determined action detection of Koga such that person passing by a building associated with the area being monitored multiple times would trigger a deterrence action. Therefore, it would have been obvious to one of ordinary skill in the art, before the effective filing date of the claimed invention, to combine the camera device of Khalili with the scene-specific machine learning model, historical data training, and package movement deterrence action including audiovisual signals of Madden, the local processor, speaker, and light of Siminoff, the integrated radar sensor and radar speed, direction, and location of movement detection of Prag, and he determined action detection of Koga to obtain the invention as specified in claim 21. Allowable Subject Matter 22. Claims 5, 10, 15, and 20 are objected to as being dependent upon a rejected base claim, but would be allowable if rewritten in independent form including all of the limitations of the base claim and any intervening claims. The following is a statement of reasons for the indication of allowable subject matter: References of record fail to specifically discloses determining movement of the object by the person to be innocuous and/or harmful and identification of the clothing of the person to classify a profession, or wherein the sex and approximate age of a person are used to discriminate whether the person is one of a child, a teenager, or an adult. Conclusion 23. 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. Inquiry 24. The prior art made of record and not relied upon is considered pertinent to applicant’s disclosure. See PTO-892. Any inquiry concerning this communication or earlier communications from the examiner should be directed to PAULO ANDRES GARCIA whose telephone number is (703)756-5493. The examiner can normally be reached Mon-Fri, 8-4:30PM ET. 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, Chan Park can be reached on (571)272-7409. 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. /PAULO ANDRES GARCIA/Examiner, Art Unit 2669 /CHAN S PARK/Supervisory Patent Examiner, Art Unit 2669