DETAILED ACTION
Notice of Pre-AIA or AIA Status
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
Status of Claims
Due to communications filed 5/14/26, the following is a final office action. Claim’s 1, 2, 5, 6, 10, 11, 14, 15, 20 are amended. Claims 4, 13 are cancelled. Claims 21-22 are new. Claims 1-3, 5-12, 14-22 are pending in this application and are rejected as follows.
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 useful improvement thereof, may obtain a patent therefor, subject to the conditions and requirements of this title,
Claims 1-3, 5-12, 14-22 are rejected under 35 U.S.C, 101 because the claimed invention is directed to a
judicial exception (I.e., a law of nature, a natural phenomenon, or an abstract idea) without significantly
more.
With regard to the present claim 1, this claim recites a judicial exception. The claim recites
receiving a video feed, detecting an image abnormality using a first machine learning component,
detecting a behavioral abnormality using a second machine learning component, and transmitting an
alert upon detecting an abnormality. These limitations collectively describe analyzing information
through video data, detecting abnormalities based on analysis, and proving a notification based on the
results. These limitations constitute "Mental Processes" (concepts performed in the human mind,
including observation, evaluation, judgment and opinion). Monitoring behavior, identifying
abnormalities, and notifying an administrator are activities that can be performed by a human observing
a video and reporting suspicious activity. Reciting that the steps of the claims are performed by using
"machine learning components" does not remove the claim from the "Mental Processes" grouping since
the claim does not recite a specific improvement to the machine learning technology but instead uses
machine learning as a tool to perform the abstract evaluation. In addition, the mere nominal recitation
of a generic computer/computer network does not take the claim out of the methods of the "Mental
Processes" grouping. Thus, the claim recites an abstract idea.
Furthermore, claim 1 is not integrated into a practical application. Claim 1 recites at least one
processor, at least one memory, at least one camera, first and second machine learning components,
and an administrator device. The claimed computer components are generic computer components
performing their ordinary functions (receiving data, processing data, comparing values to thresholds,
and transmitting notifications. Simply implementing the abstract idea on a generic computer is not a
practical application of the abstract idea. In addition, the limitations beyond the abstract idea include
"at least one processor", "at least one memory", "at least one camera", and "an administrative device"
are recited at a high level of generality and represent generic computer components performing their
conventional functions (receiving data, processing data and transmitting data), without improving the
functioning of a computer or another technology, reciting a particular machine learning architecture or
specific technical improvement in video processing, or apply the abstract idea in a meaningful way
beyond linking it to generic computer implementation. The generic components of the claim are used as
tools to perform the abstract idea of the invention, and therefore the judicial exception is not integrated
into a practical application.
Finally, claim 1 does not recite an inventive concept. The additional elements amount to a
generic processor and memory, generic camera input, generic data processing, and generic data
transmission. Thus, even when viewed as a whole, nothing in the claim adds significantly more (i.e., an
inventive concept) to the abstract idea. Specifically, the claim recites a processor, a memory, a camera
and a transmission of an alert, which are well-understood, routine and conventional components
performing conventional functions of data reception, analysis and notification. With regard to the first
and second machine learning components, these do not provide an inventive concept because there is
no specification of any unconventional training technique, model, structure or technical improvement.
In fact, using machine learning to detect abnormalities in a video is a conventional data analysis
technique. When considered individually and as an ordered combination, the additional elements
merely implement the abstract idea using generic computing functionality. Claim 1 therefore does not
amount to significantly more than the abstract idea itself. The claim is ineligible.
Dependent claim 2-9 depend from independent claim 1 and are therefore rejected under 35
USC 101 based on their dependencies on independent claim 1.
Independent claim 10 recites limitations similar to those of independent claim 1 and is therefore
rejected for similar reasons.
Dependent claim 11-19 depend from independent claim 10 and are therefore rejected under 35
USC 101 based on their dependencies on independent claim 10.
Independent claim 20 recites limitations similar to those of independent claim 1 and is therefore
rejected for similar reasons.
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.
Claim(s) 1-6 and 8-20 is/are rejected under 35 U.S.C. 103 as being unpatentable over Chun et al (US 20250384735 A1), and further in view of Kim et al (20230005269 A1).
As per claim 1, Chun et al discloses:
receive at least one video feed from at least one camera, ([10506] 1. Real-Time Video Analysis: The
Integrated Camera System continuously streams high-definition video footage into the Buffer Recording
System and the Al-Based Anomaly Detection Module);
detect, using a first proctoring machine learning component, at least one user of the self-enrollment system and at least one image abnormality in a frame of the at least one video feed, ([0031] A notable security innovation in this system is the integration of multiple sensors that monitor environmental conditions and player interactions in real time; [0354] In at least one embodiment, the Tournament and Bonus Server Component(s) 530 are configured to manage the entire lifecycle of multiplayer tournaments, including player registration; [6931] The Dice Shaker Gaming System (Electro-Mechanical Gaming Terminal) serves as the primary gaming interface where players register for tournaments, place structured bets, and track their ranking progress. The system integrates with the tournament management module to allow players to participate in competitive dice-based gameplay; [10506] The Al module performs real-time image processing using machine learning algorithms trained to identify anomaly signatures, such as unauthorized shaking or tampering behaviors. Each video frame is analyzed for motion vectors, behavioral patterns, and environmental disturbances. The Al module processes thousands of frames per second, evaluating player movements, interactions with the DSG system, and environmental factors to differentiate between normal and suspicious behavior;
detect, using a second proctoring machine learning component, at least one behavioral abnormality associate with the at least one user in a segment of the video feed, (([0031] A notable security innovation in this system is the integration of multiple sensors that monitor environmental conditions and player interactions in real time; [10506] The AI module performs real-time image processing using machine learning algorithms trained to identify anomaly signatures, such as unauthorized shaking or tampering behaviors. Each video frame is analyzed for motion vectors, behavioral patterns, and environmental disturbances. The Al module processes thousands of frames per second, evaluating player movements, interactions with the DSG system, and environmental factors to differentiate between normal and suspicious behavior; [10569] AI-powered player behavior analysis may further enhance gaming experiences by detecting wagering patterns, preferences, and engagement trends. A neural network-based recommendation engine may analyze individual player histories to suggest game modes, betting options, or promotional offers tailored to each user. This system may also detect potential problem gambling behaviors, triggering responsible gaming interventions such as automated alerts, wagering limits, or temporary session restrictions.);
transmit an alert to an administrator device upon detecting at least one abnormality, [10510] This
process involves frame-by-frame analysis of the dice's final resting position before and after the
detected anomaly. The Al module computes deviation scores to quantify the potential impact on the
outcome, triggering a flag if discrepancies are detected. [10511] 6. Security Alert Generation and Data
Compilation: Simultaneously, the Security Alert System compiles all processed data into a structured
security alert report. This report includes the saved video segment, metadata, Al validation findings, and
an initial risk assessment This data processing step ensures that the alert is both comprehensive and
securely packaged for immediate delivery to security personnel).
With regard to the following limitation:
receiving an actual frame from the at least one video feed; determining a difference value between the actual frame and the predicted frame; and comparing the difference value to a threshold value.
Chun et al discloses the following:
([0660] The system operates by continuously monitoring dice roll outcomes, comparing them against expected probability distributions, and flagging any statistical irregularities. If an anomaly is detected-such as repeated numbers appearing beyond expected thresholds-the system automatically triggers an alert; [0626] AI Image Recognition Module (1511): The AI image recognition module (1511) is an advanced computer vision system that analyzes dice roll outcomes, verifies randomness, and ensures accurate game results... These cameras capture multiple frames per second, allowing the AI system to analyze dice motion, final resting positions, and potential irregularities...;[0628] Some embodiments incorporate AI-driven predictive adjustments that modify rolling dynamics based on past outcomes, ensuring fair randomness over extended play sessions; [0654] The system operates by continuously analyzing dice roll data over extended periods, identifying trends that deviate from expected statistical distributions. If roll sequences show recurring anomalies—such as a specific number appearing more frequently than probability models predict—the system generates alerts for manual review...The module may also provide predictive maintenance insights, helping casino operators address potential mechanical issues before they impact fairness.
Chun et al does not disclose:
generating a predicted frame of the at least one video feed based on a previous frame;
receiving an actual frame from the at least one video feed; determining a difference value between the actual frame and the predicted frame; and comparing the difference value to a threshold value.
However, Kim et al discloses: [0006] According to an aspect of an exemplary embodiment, a method of detecting an abnormal event in a series of temporally successive images includes: generating a predicted current frame based on a previous frame temporally ahead of an actual current frame and a subsequent frame temporally behind the actual current frame; calculating an anomaly score indicating a difference between the predicted current frame and the actual current frame; and determining that an abnormality is included in the actual current frame when the anomaly score satisfies a predetermined condition.
It would have been obvious to one of ordinary skill in the art at the time the invention was made to include the above limitations as taught by Kim et al in the systems of Chun et al, since the claimed invention is merely a combination of old elements, and in the combination each element merely would have performed the same function as it did separately, and one of ordinary skill in the art would have recognized that the results of the combination were predictable.
Chun does not specifically disclose a proctoring machine, however,
As per claim 2, Chun et al discloses:
wherein the detecting at least one image abnormality is based on comparing at least one measured
value of a detected behavior over a time period to at least one threshold value, ([0660], The system operates by continuously monitoring dice roll outcomes, comparing them against expected probability distributions, and flagging any statistical irregularities. If an anomaly is detected-such as repeated numbers appearing beyond expected thresholds the system automatically triggers an alert; [10510] This process involves frame-by-frame analysis of the dice's final resting position before and after the detected anomaly. The module compares dice position coordinates, motion signatures, and environmental factors to assess if unauthorized player actions (like shaking the DSG system) altered the dice outcome. The validation process also cross-references the saved game state data, including the final dice result and timestamps, to ensure consistency. The Al module computes deviation scores to quantify the potential impact on the outcome, triggering a flag if discrepancies are detected; [0057] In at least one embodiment, the system further comprises a camera-based anomaly detection and response system including at least one camera and a memory buffer system, the camera-based anomaly detection and response system being configured to cause the at least one processor to execute instructions for continuously recording video data of gameplay activities over a first predetermined time interval using the memory buffer system and cyclically overwriting older data, automatically saving a video clip from the memory buffer system that brackets a timestamp of a detected anomaly event, generating an alert message in response to detecting the anomaly event);
As per claim 3, Chun et al discloses:
wherein the at least one measured value includes at least one of: a facial detection, an object detection,
a hand detection, an emotion detection, a pose detection, or an eye gaze detection, ([0172] The
transparent housing (451) encloses the electro-mechanical dice RNG mechanism (495), creating a secure
and visible rolling chamber that allows for real-time observation of the dice rolling process. This
enclosure is specifically designed to maintain the integrity of each dice roll by preventing external
interference while providing a clear, unobstructed view of the game mechanics. The housing integrates
seamlessly with security sensor systems, including proximity sensors that detect unauthorized objects or
hands attempting to interact with the dice shaker; [0510] Sensor(s)/Camera(s) 950 may be configured or
designed to detect and capture external data, events, and/or conditions including, for example,
biometric information (e.g., facial images, facial features).
As per claim 5, Chun et al discloses:
wherein the second proctoring machine learning component comprises a long-short term memory network, ([9468] If a multi-level rolling sequence occurs, the system: [9469] Transmits intermediate roll results; [9470] Updates player balances dynamically as rolls progress. [9471] Ensures that jackpot conditions or progressive wagers are reflected in real time.);
and wherein detecting at least one behavioral abnormality further includes referencing at least one
previous segment of the video feed, ([0626] These cameras capture multiple frames per second,
allowing the Al system to analyze dice motion, final resting positions, and potential irregularities The
system is designed to integrate with external compliance auditing platforms, where regulators may
review image logs of past game events. Al model updates may be performed regularly to improve
detection accuracy and fraud prevention measures. The benefits of the Al image recognition module
include enhanced roll accuracy, reduction in fraudulent activities, and regulatory compliance; [1182] The
AI recognition system identifies the numeric values of the dice. [1183] The values are cross-verified
against previous frame captures to ensure accuracy).
As per claim 6, Chun et al discloses:
wherein the at least one processor is further configured to train the second proctoring machine learning
component on at least one training data including one or more example enrollment processes, ([1892]
To prevent false positives caused by legitimate player movements, the system employs Al-driven fraud
detection that continuously learns from gameplay behavior. If a player moves near the machine or
interacts with the player terminal in a way that triggers an alert, the Al module distinguishes between
normal interaction (such as adjusting the viewing angle or leaning forward to place a bet) and fraudulent
behavior (such as attempting to physically influence the dice roll). The AI system is trained to adapt and
improve over time, minimizing unnecessary gameplay interruptions and ensuring real-time fraud
detection without compromising the player experience);
where the training further includes altering the at least one training data with at least one of:
downsizing, gray scaling, manual data review, and consecutive frame selection, ([1893] The system logs
every action taken during gameplay, including all security events, sensor data, and AI classifications,
ensuring that tampering attempts are fully documented. The logs are stored in a tamper-proof
database, providing a transparent audit trail for casino operators, regulatory authorities, and security
personnel. This ensures that all security responses are traceable, and that any suspicious behavior may
be audited and investigated. In case of a major security incident, such as coordinated tampering across
multiple machines, the system may automatically alert casino-wide security to allow swift intervention
and minimize the impact of fraud; [0640] Regulatory Audit & Compliance System (1527): The regulatory
audit & compliance system (1527) is responsible for ensuring that all aspects of the wager-based gaming
system adhere to legal and industry standards. This system continuously monitors game operations,
records gameplay events, and generates compliance reports for regulatory authorities Automated
algorithms review this data for compliance with gaming regulations, and if anomalies are detected, the
system generates alerts for manual review).
As per claim 8, Chun et al discloses:
store one or more unaddressed alerts into a queue, ([10559] 7. Security Interface Reset: The Security
Personnel Interface is updated to reflect the incident's closure, removing the active alert from the
immediate review queue and filing it into the resolved incidents archive; [this step shows the alert is
stored in a queue);
store, using a database, at least one abnormality information about at least one detected abnormality as
stored abnormality data; and send a record of stored abnormality data to the administrator device,
([1893] The system logs every action taken during gameplay, including all security events, sensor data,
and Al classifications, ensuring that tampering attempts are fully documented. The logs are stored in a
tamper-proof database, providing a transparent audit trail for casino operators, regulatory authorities,
and security personnel. This ensures that all security responses are traceable, and that any suspicious
behavior may be audited and investigated. In case of a major security incident, such as coordinated
tampering across multiple machines, the system may automatically alert casino-wide security to allow
swift intervention and minimize the impact of fraud).
As per claim 9, wherein the alert includes abnormality information associated with the at least one
abnormality and the alert is displayed on the administrator device, ([1350] If a tampering event is
detected, the system: [1351] Locks the dice shaker mechanism. [1352] Displays a security alert. [1353]
Sends a compliance notification to casino staff; [1414] If the system detects unauthorized movement, it:
[1415] Stops the dice shaker unit immediately. [1416] Prevents payout calculations until the issue is
resolved. [1417] Displays an on-screen message alerting players and casino staff; [1459] The screen
displays a message to the player informing them that unauthorized movement has been detected and
that the game has been temporarily suspended for security verification. Simultaneously, casino security
personnel receive an automated alert containing real-time sensor logs, impact force data, and a
timestamp of the event. The compliance monitoring system logs the incident in the casino's regulatory
database, ensuring that all security-related activities are fully recorded for auditing.).
As per claim 10, this claim recites limitations similar to those disclosed in independent claim 1 and is
therefore rejected for similar reasons.
As per claim 11:
wherein the detecting at least one image abnormality is based on comparing at least one measured
value of a detected behavior over a time period to at least one threshold value.
Please see the rejection for claim 2.
As per claim 12:
wherein the at least one measured value includes at least one of: a facial detection, an object detection,
a hand detection, an emotion detection, a pose detection, or an eye gaze detection.
Please see the rejection for claim 3.
As per claim 14:
Wherein the second proctoring machine learning component comprises a long-short term memory
network, and wherein detecting at least one behavioral abnormality further includes referencing at least one previous segment of the video feed.
Please see the rejection for claim 5.
As per claim 15:
Wherein the at least one processor is further configured to train the second proctoring machine learning
component on at least one training data including one or more example enrollment processes, where
the training further includes altering the at least one training data with at least one of: downsizing, gray
scaling, manual data review, and consecutive frame selection.
Please see the rejection for claim 6.
As per claim 16:
wherein the at least one processor is further configured to:
display, using an administrator device, the at least one video feed, wherein displaying the at least one
video feed further includes:
drawing a bounding box around an area in which an abnormality is detected;
generating text defining a bounding box based on a type of the abnormality; and
displaying the bounding box and text using the administrator device.
Please see the rejection for claim 7.
As per claim 17:
Wherein the at least one processor is further configured to:
store one or more unaddressed alerts into a queue; and
store, using a database, at least one abnormality information about at least
one detected abnormality as stored abnormality data.
Please see the rejection for claim 8.
As per claim 18:
wherein the at least one processor is further configured to:
send a record of stored abnormality data to the administrator device.
Please see the rejection for claim 8.
As per claim 19:
wherein the alert includes abnormality information associated with the at least one abnormality and the
alert is displayed on the administrator device.
Please see the rejection for claim 9.
As per claim 20, this claim recites limitations similar to those recited in independent claim 1, and is
therefore rejected for similar reasons.
Claim(s) 7 is/are rejected under 35 U.S.C. 103 as being unpatentable over Chun et al (US
20250384735 A1), and further in view of Kim et al (20230005269 A1), and further in view of (TH 106597 B).
As per claim 7, Chun et al does not disclose the following limitations, however, (TH 106597 B)
discloses:
display, using an administrator device, the at least one video feed, wherein
displaying the at least one video feed further includes:
drawing a bounding box around an area in which an abnormality is
detected; generating text defining a bounding box based on a type of the
abnormality; and
However, (TH 106597 B), discloses: returning to Figure 3. In the non-matching object identification step,
the non-matching object identification unit 312 feeds multiple images acquired by the image acquisition
unit 310 This makes it possible for the operator to know exactly where each incompatible object is
located. The location of the incompatible objects is also displayed sup., Incompatible objects may be
represented by surrounding the area where the objects are located with a frame, etc. Incompatible
objects may also be represented by representation of a predefined bounding box for each incompatible
object, or a display of the incompatible object type with text, etc., may also be used. In addition, the
probability of an incompatible object being present may be represented by displaying the probability of
the incompatible object being present in the area enclosed by a numerical bounding box on the display);
However, (TH 106597 B) discloses: displaying the bounding box and text using the administrator device,
(The type and location of incompatible objects and their probability are identified. In addition, the
location of incompatible objects when detecting incompatible objects is also identified, which makes it
possible for operators to visually confirm the location of incompatible objects in the image
immediately. The scrap metal as it is being transported is displayed on the display and the location,
probability, and type of incompatible objects. as it is being transported are also displayed The location
of the incompatible objects is also displayed. sup., Incompatible objects may be represented by
surrounding the area where the objects are located with a frame, etc. Incompatible objects may also be
represented by changing their color, A shape, or similar representation of a predefined bounding box for
each incompatible object, or a display of the incompatible object type with text, etc.).
It would have been obvious to one of ordinary skill in the art at the time the invention was made to
include the above limitations as taught by (TH 106597 B) in the systems of Chun et al, since the claimed
invention is merely a combination of old elements, and in the combination each element merely would
have performed the same function as it did separately, and one of ordinary skill in the art would have
recognized that the results of the combination were predictable.
Claim(s) 21, 22 is/are rejected under 35 U.S.C. 103 as being unpatentable over Chun et al (US
20250384735 A1), and further in view of Kim et al (20230005269 A1), and further in view of (KR 102738912 B1).
As per claim 21, Chun et al does not disclose the following, however, (KR 102738912 B1) discloses:
detecting using the first proctoring machine learning component, facial landmarks associated with at least one detected face; and determining an emotion based on the facial landmarks, (KR 102738912 B1):
In relation to emotional state, the system can perform image analysis to detect the emotional state of the patient. For example, the system can determine the emotional state of the patient by performing facial feature detection on image data from the patient.; At step 1515, the system can perform face detection within the image data. The system identifies the patient's face within the image data received at step 1510. At step 1520, the system can detect locations of facial landmarks, such as eyebrows, nose, lips, etc., within the image data. In one embodiment, the system performs facial landmark detection using any of a variety of methods that can combine HOG descriptors and other low-level vision features with advanced machine learning and classification techniques. At step 1525, the system can perform local facial region extraction using these facial landmarks. The system can analyze the local facial regions for intensity variations associated with the facial landmarks).
It would have been obvious to one of ordinary skill in the art at the time the invention was made to
include the above limitations as taught by (KR 102738912 B1) in the systems of Chun et al, since the claimed invention is merely a combination of old elements, and in the combination each element merely would have performed the same function as it did separately, and one of ordinary skill in the art would have recognized that the results of the combination were predictable.
As per claim 21, Chun et al does not disclose the following, however, (KR 102738912 B1) discloses:
wherein the detecting at least one behavioral abnormality is based on: detecting using the first proctoring machine learning component, facial landmarks associated with at least one detected face; and determining an emotion based on the facial landmarks, ((KR 102738912 B1): In relation to emotional state, the system can perform image analysis to detect the emotional state of the patient. For example, the system can determine the emotional state of the patient by performing facial feature detection on image data from the patient.; At step 1515, the system can perform face detection within the image data. The system identifies the patient's face within the image data received at step 1510. At step 1520, the system can detect locations of facial landmarks, such as eyebrows, nose, lips, etc., within the image data. In one embodiment, the system performs facial landmark detection using any of a variety of methods that can combine HOG descriptors and other low-level vision features with advanced machine learning and classification techniques. At step 1525, the system can perform local facial region extraction using these facial landmarks. The system can analyze the local facial regions for intensity variations associated with the facial landmarks).
It would have been obvious to one of ordinary skill in the art at the time the invention was made to
include the above limitations as taught by (KR 102738912 B1) in the systems of Chun et al, since the claimed invention is merely a combination of old elements, and in the combination each element merely would have performed the same function as it did separately, and one of ordinary skill in the art would have recognized that the results of the combination were predictable.
Response to Arguments
Applicant’s arguments, see the, filed 5/14/26, with respect to “Claim Interpretation” rejection have been fully considered and are persuasive. The “Claim Interpretation” rejection of claim 1 has been withdrawn.
Applicant's arguments filed 5/14/26 have been fully considered but they are not persuasive.
With regard to the 101 rejection, Applicant argues that the pending claims are not directed to an abstract idea and specifically discloses that with regard to step 2A, prong 1, the claims recite computer-specific and machine-learning—specific activity relating to the manipulation of image data from a live video fee. However Examiner respectfully disagrees. Although the claims recite receiving a video feed, detecting image and behavioral abnormalities using machine learning components, generating a predicted frame, determining a difference value between a predicted frame and an actual frame, comparing the difference value between a predicted frame and an actual frame, comparing the difference value to a threshold, and transmitting an alert, these additional elements are recited at a high level of generality and are used to implement the analysis of human behavior. The claimed processor, memory, camera, and machine learning components are invoked as tools for performing the recited analysis rather than an improvement to the functioning of the computer itself or another technology. The claim does not recite a specific machine-learning architecture or a particular improvement to image-processing technology. For example, while the specification may describe technical implementation details such as encoder-decoder networks, LTSM cells, latent feature extraction, or temporal feature preservation, those features are not affirmatively recited in the claims. The eligibility analysis is based on the claimed invention. Accordingly, the claims are directed to collecting information from a video feed, analyzing that information to detect abnormalities, and providing an alert, a mental process. The additional elements do not integrate the exception into a practical application because they merely use generic computing components to perform the recited analysis. However,
With regard to Applicant’s arguments with regard to Step 2A, prong 2, Applicant contends that the claims are integrated into a practical application because they recite computer-specific and machine-learning specific operations for manipulating image data from a live video feed. Applicant further argues that the claims require generating a predicted frame based on a previous frame , comparing the predicted frame to an actual frame and determining behavioral abnormalities based on the comparison. However, these limitations are recited at a high level of generality and are used to perform the analysis of video data to identify abnormal behavior. The claims do not recite a specific improvement to computer functionality or image-processing technology, nor do they recite a particular machine-learning architecture that improves the operation of a computer or another technology. While the specification describes embodiments utilizing an encoder, LSTM cells configured to preserve temporal information from pervious frames, and a decoder configured to reconstruct input data based on learned latent features, these features are not affirmatively recited in the pending claims. Likewise, the specification’s discussion of processing spatial and temporal information across sequences of frames describes exemplary implementations but does not limit the scope of the claims. The eligibility analysis under 35 USC 101 is based on the claimed invention rather than the unclaimed invention. Accordingly, the additional elements do not integrate the recited judicial exception into a practical application. Instead, they employ generic computer components and recite machine-learning components as tools to collect, analyze and evaluate video information in order to detect abnormalities and transmit an alert. Such use of computer technology to perform the recited analysis does not constitute an improvement to computer functionality or another technology. Therefore the claims remain directed to a judicial exception without integrating that exception into a practical application.
With regard to Applicant’s argument for Step 2B, and that the present claims satisfy the Step 2B requirement, because the claims include specific machine-learning operations for processing live video data, including generating a predicted frame, receiving an actual frame, determining a difference value between the predicted frame and the actual frame, comparing the difference value to a threshold, and transmitting an alert. However, Examiner respectfully disagrees. The additional claim elements, considered individually and as an ordered combination, do not amount to significantly more than the judicial exception. The claimed processor, memory, camera, machine-learning components, and administrative device are recited at a high level of generality and perform their ordinary and expected functions of receiving data, processing data, analyzing data, comparing data, and transmitting results. The claims do not recite any specialized hardware or any specific improvement to the functioning of a computer or other technology. Further, although the claims recite generating a predicted frame, determining a difference value, and comparing the difference value to a threshold, the claims do not recite a particular algorithm or machine-learning architecture that is asserted to improve computer functionality. Rather, these limitations are used as part of the analysis of video data to detect abnormal behavior and provide a notification. Accordingly, these additional elements merely apply the judicial exception using generic computer technology and do not impose a meaningful limit on the judicial exception. Therefore, when viewed individually and as an ordered combination, the additional elements do not amount to significantly more than the judicial exception. Accordingly, the rejection under 35 U.S.C. § 101 is maintained.
Applicant’s arguments, see arguments/remarks, filed 5/14/26, with respect to the rejection(s) of claim(s) 1-6 and 8-20 as being rejected under Chun et al (US 20250384735 A1), have been fully considered and are persuasive. Therefore, the rejection has been withdrawn. However, upon further consideration, a new ground(s) of rejection is made in view of and further in view of Kim et al (20230005269 A1).
Similarly, claim(s) is now rejected under Chun et al (US 20250384735 A1), and further in view of Kim et al (20230005269 A1), and further in view of (TH 106597 B).
Similarly, claim(s) 21, 22 is/are now rejected under Chun et al (US20250384735 A1), and further in view of Kim et al (20230005269 A1), and further in view of (KR 102738912 B1).
Conclusion
THIS ACTION IS MADE FINAL. Applicant is reminded of the extension of time policy as set forth in 37 CFR 1.136(a).
A shortened statutory period for reply to this final action is set to expire THREE MONTHS from the mailing date of this action. In the event a first reply is filed within TWO MONTHS of the mailing date of this final action and the advisory action is not mailed until after the end of the THREE-MONTH shortened statutory period, then the shortened statutory period will expire on the date the advisory action is mailed, and any nonprovisional extension fee (37 CFR 1.17(a)) pursuant to 37 CFR 1.136(a) will be calculated from the mailing date of the advisory action. In no event, however, will the statutory period for reply expire later than SIX MONTHS from the mailing date of this final action.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to Akiba Robinson whose telephone number is 571-272-6734 and email is Akiba.Robinsonboyce@USPTO.gov. The examiner can normally be reached on Monday-Thursday 6:30am-4:30pm.
If attempts to reach the Examiner by telephone are unsuccessful, the Examiner's supervisor, Nathan Uber can be reached on 571-270-3923. The fax phone number for the organization where this application or proceeding is assigned is 571-273-8300.
Any inquiry of a general nature or relating to the status of this application or proceeding should be directed to the receptionist whose telephone number is (703) 305-3900.
July 9, 2026
/AKIBA K ROBINSON/Primary Examiner, Art Unit 3626
Prosecution Insights