DETAILED ACTION
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
This office action is in response to communication/amendment filed on 08/25/2025.
Status of claims in the instant application:
Claims 1-25 and 27-29 are pending.
Claim 26 has been canceled.
No new claim has been added.
Claims 1-25 and 27-29 have been amended.
Response to Arguments
Applicant’s arguments, see page [9] of the remarks filed on 08/25/2025 with respect to objection to claims, have been fully considered in view of the claim amendments and are persuasive. Therefore, the claim objections have been withdrawn.
Applicant’s arguments, see page [9] of the remarks filed on 08/25/2025 with respect to rejection of claims under 35 USC 112(b), have been fully considered in view of claim amendments and are persuasive. Therefore, the claim rejections have been withdrawn.
Applicant’s arguments, see page [9-11] of the remarks filed on 08/25/2025 with respect to rejection of claims under 35 USC 103, have been fully considered in view of the claims amendments and but they are not persuasive. Therefore, the claim rejections have been maintained in this office action. Furthermore, Applicant’s claim amendments have rendered new grounds for rejection.
Therefore, the Applicant is directed to Examiner’s response below.
Applicant states, see page [9-10] of the remarks, with respect to amended independent claim 1:
“However, Applicant respectfully submits that the cited portions of Shepherd do not teach or suggest “a behavior pattern label hierarchy that comprises a plurality of behavior pattern labels that are hierarchically related based on a respective skill level, wherein the skill level is beginner level, intermediate level, or expert level,” as recited in amended claim 1. Rather, Shepherd describes generating machine-learned trust scores and categorizing players based on likelihood to exhibit certain behaviors (e.g., cheating, griefing, vulgar language, game abandonment) for purposes of matchmaking. Shepherd at [0043]. Shepherd does not disclose or suggest a behavior pattern label hierarchy in which multiple behavior pattern labels are organized in a hierarchical relationship explicitly based on discrete skill levels, nor does Shepherd describe structuring such labels within a hierarchy.
In response, Examiner respectfully disagrees with Applicant’s characterization of Shepherd prior art not disclosing Applicant’s claimed feature, as noted above.
Examiner highlights the following from previously cited Para [0043] of Shepherd prior art:
“Shepherd, Para [0043]: As yet another example, the trained machine learning model(s) 216 may be configured to output a trust score 118 that relates to a probability of a player behaving, or not behaving, in accordance with a “high-skill” behavior. In this manner, the scoring can be used to identify highly-skilled players, or novice players, from a set of players. This may be useful to prevent situations where experienced gamers create new user accounts pretending to be a player of a novice skill level just so that they can play with amateur players. Accordingly, the players matched together in the first match 218(1) may be those who are likely (as determined from the machine-learned scores 118) to behave in accordance with a particular “bad” behavior, while the players matched together in other matches, such as the second match 218(2) may be those who are unlikely to behave in accordance with the particular “bad” behavior.”
Examiner also reproduces the following from published instant application (US 20240179167 A1):
“US 20240179167 A1, Para [0032]: Optionally, the behaviour pattern label and the behaviour pattern label hierarchies relate to a skilled of the user of the electronic device. Preferably, the behaviour pattern label hierarchies comprise beginner, intermediate, and expert. Optionally, the likelihood of an anomaly occurring relates to one of: anomaly, no anomaly, or suspicion of anomaly. Optionally, an anomaly is indicative of the user cheating at a game being played on the electronic device.”
Applicant’s claimed feature of skill level[s] is used to detect anomalous behavior. Shepherd, Para [0043] also discloses the use of skill level[s] to detect cheating (anomaly) by player[s] in video game, disclosing Applicant’s claimed feature.
Examiner further notes that Applicant's arguments fail to comply with 37 CFR 1.111(b) because they amount to a general allegation that the claims define a patentable invention without specifically pointing out how the language of the claims patentably distinguishes them from the references.
Applicant’s remaining arguments regarding claims 27 and 28 refer to the previous arguments for claim 1 that the Examiner has already addressed above.
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 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.
Claims 1-16, 18-25 and 27-28 are rejected under 35 U.S.C. 103 as being unpatentable over Pub. No.: US 2021/0232907 A1 to Pardeshi et al. (hereinafter “Pardeshi”) in view of Pub. No.: US 2021/0038979 A1 to Bleasdale-Shepherd et al. (hereinafter “Shepherd”).
Regarding Claim 1. Pardeshi discloses A computer implemented method for autonomous anomaly device operation detection (Pardeshi, Abstract, Para [0055-0067]: … Apparatuses, systems, and techniques to detect cheating, manipulation, or unfair advantages. In at least one embodiment, cheating is determined using a reconstruction probability inferred, using one or more neural networks, for video data for a player of a game …), the method comprising the steps:
receiving behavior data, wherein the behavior data is indicative of a user's inputs to an electronic device (Pardeshi, FIG. 2A, FIG. 4-5, Para [0053, 0070] … In at least one embodiment, client devices 204 are used by players to play sessions (or other portions of levels) of a specific game application, as may be hosted locally, peer-to-peer, or on a number of game servers 202. In at least one embodiment, client devices 204 may be any appropriate electronic devices enabling players to participate in a particular gaming session, as may include desktop computers, notebook computers, smartphones, tablet computers, gaming consoles (portable or otherwise), and set-top boxes. In at least one embodiment, player devices 204 communicate over at least one network 216 with at least one game server 202 in order to participate in a game session … In at least one embodiment, a process 400 for determining cheating can be used as illustrated in FIG. 4. In at least one embodiment, a stream of media data is received 402 corresponding to a gaming session for a player. In at least one embodiment, this can include video data generated by a gaming application executing on a client device. In at least one embodiment, segments of this media data can be provided 404 as input to a cheating determination system …),
determining a behavior pattern label and an indication to the likelihood of an anomaly occurring based on the received behavior data (Pardeshi, Para [0070-0072]: … In at least one embodiment, a segmentation module can be used to select segments, such as by using a random or determined selection algorithm. In at least one embodiment, specific segments may be received that correspond to highlights or certain types of occurrences in a game … In at least one embodiment, these segments are analyzed by a reconstruction module to calculate 406, using at least one neural network, a reconstruction probability for these segments, where individual reconstruction probabilities are inferred for individual frames of those segments. In at least one embodiment, a labeling module may also be used to determine 408, using one or more neural networks, labels for events or occurrences in those segments, where those events or occurrences may correspond to objects, scenes, or actions. In at least one embodiment, these labels may correspond to evidence of cheating learned for one or more games, types of games, or levels … In at least one embodiment, a determined reconstruction probability can be provided 410, along with any determined labels, to a decision server. In at least one embodiment, this decision sever can determine 412, using at least one neural network, whether one or more anomalies determined in analyzed segments corresponds to cheating …), [wherein the behavior pattern label belongs to behavior pattern label hierarchy that comprises a plurality of behavior pattern labels that are hierarchically related based on a respective skill level, wherein the skill level is beginner level, intermediate level, or expert level],
However, Pardeshi does not explicitly teach, but Shepherd from same or similar field of endeavor teaches:
“wherein the behavior pattern label belongs to behavior pattern label hierarchy that comprises a plurality of behavior pattern labels that are hierarchically related based on a respective skill level, wherein the skill level is beginner level, intermediate level, or expert level (Shepherd, Para [0036, 0038, 0043, 0063-0064]: … In some embodiments, other factors (e.g., skill level, geographic region, etc.) are considered in the matchmaking process, which may cause further breakdowns and/or subdivisions of players into a fewer or greater number of matches 218 … FIG. 2 illustrates examples of other behaviors, besides cheating, which can be used as a basis for player matchmaking. For example, the trained machine learning model(s) 216 may be configured to output a trust score 118 that relates to the probability of a player behaving, or not behaving, in accordance with a game-abandonment behavior (e.g., by abandoning (or exiting) the video game in the middle of a match). Abandoning a game is a behavior that tends to ruin the gameplay experience for non-abandoning players, much like cheating … Oftentimes, multiplayer video games allow for players to engage in chat sessions or other social networking communications that are visible to the other players in the video game 110, and when a player uses vulgar language (e.g., curse words, offensive language, etc.), it can ruin the gameplay experience for players who do not use vulgar language. As yet another example, the trained machine learning model(s) 216 may be configured to output a trust score 118 that relates to a probability of a player behaving, or not behaving, in accordance with a “high-skill” behavior. In this manner, the scoring can be used to identify highly-skilled players, or novice players, from a set of players. This may be useful to prevent situations where experienced gamers create new user accounts pretending to be a player of a novice skill level just so that they can play with amateur players. Accordingly, the players matched together in the first match 218(1) may be those who are likely (as determined from the machine-learned scores 118) to behave in accordance with a particular “bad” behavior, while the players matched together in other matches, such as the second match 218(2) may be those who are unlikely to behave in accordance with the particular “bad” behavior …)”
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 teachings of Shepherd into the teachings of Pardeshi, because it discloses that, “the scoring can be used to identify highly-skilled players, or novice players, from a set of players. This may be useful to prevent situations where experienced gamers create new user accounts pretending to be a player of a novice skill level just so that they can play with amateur players (Shepherd, Para [0043])”.
Pardeshi further discloses:
“providing the behavior pattern label and the likelihood of an anomaly occurring (Pardeshi, Para [0072-0073]: … In at least one embodiment, a determined reconstruction probability can be provided 410, along with any determined labels, to a decision server. In at least one embodiment, this decision sever can determine 412, using at least one neural network, whether one or more anomalies determined in analyzed segments corresponds to cheating … In at least one embodiment, it can be determined 506 whether that player is cheating in that game based at least in part upon this reconstruction probability …)”.
Regarding Claim 2. The combination of Pardeshi-Shepherd discloses the method according to claim 1, Pardeshi further discloses, “wherein determining the behavior pattern label and the indication to the likelihood of an anomaly occurring comprises:
processing the behavior data using a behavior machine learning model (Pardeshi Para [0055-0058]: … In at least one embodiment, cheating detection for applications such as video games can utilize semi-supervised machine learning techniques to analyze gameplay. In at least one embodiment, such a system can accept as input a gameplay video clip or segment, and can process that video data for anomalies. In at least one embodiment, anomaly detection is performed using one or more Variational AutoEncoders (VAEs). In at least one embodiment, labels are generated for this video data using one or more machine learning-based techniques, which can help to determine to a type cheating being performed …),
wherein the behavior pattern label is based on the output of the behavior machine learning model (Pardeshi Para [0070-0072]: … In at least one embodiment, a labeling module may also be used to determine 408, using one or more neural networks, labels for events or occurrences in those segments, where those events or occurrences may correspond to objects, scenes, or actions. In at least one embodiment, these labels may correspond to evidence of cheating learned for one or more games, types of games, or levels … In at least one embodiment, a determined reconstruction probability can be provided 410, along with any determined labels, to a decision server. In at least one embodiment, this decision sever can determine 412, using at least one neural network, whether one or more anomalies determined in analyzed segments corresponds to cheating ).
Regarding Claim 3. The combination of Pardeshi-Shepherd discloses the method according to claim 2, Pardeshi further discloses, “wherein the behavior machine learning model is an artificial neural network (Pardeshi, Para [0001, 0255]: … At least one embodiment pertains to processing resources used to perform and facilitate artificial intelligence. For example, at least one embodiment pertains to processors or computing systems used to train neural networks according to various novel techniques described herein … In at least one embodiment, neurons 2202 in second layer 2212 may fan out to neurons 2202 in multiple other layers, including to neurons 2202 in (same) second layer 2212. In at least one embodiment, second layer 2212 may be referred to as a “recurrent layer.” In at least one embodiment, neuromorphic processor 2200 may include, without limitation, any suitable combination of recurrent layers and feed-forward layers, including, without limitation, both sparsely connected feed-forward layers and fully connected feed-forward layers …).”
Regarding Claim 4. The combination of Pardeshi-Shepherd discloses the method according to claim 2, Pardeshi further discloses, “further comprising the step of:
receiving user perspective data, wherein the user perspective data is indicative of a visual component of a scene being displayed by the electronic device, and wherein the step of determining the behavior pattern label and the indication to the likelihood of an anomaly occurring (Pardeshi, Para [0006]: … FIG. 3 illustrates gameplay elements such as objects, scenes, and actions that can be identified and labeled for use in a cheating determination, according to at least one embodiment …) further comprises:
processing the user perspective data using a user perspective machine learning model, and wherein the behavior pattern label and the indication to the likelihood of an anomaly occurring are additionally based on the output of the user perspective machine learning model (Pardeshi, Para [0006, 0050]: … In at least one embodiment, a cheating player can utilize a technique that performs actions on behalf of that player, without requiring some or any player input. In at least one embodiment, this can include code taking an action, such as to cause a golf club 134 to be swung to hit a golf ball 132 towards a hole as illustrated in video frame 130 of FIG. 1B. In at least one embodiment, a position of a cursor 136 used for player input does not match an expected position of that cursor to provide input that matches this swing of golf club 134. In at least one embodiment, this can be measured as a delta between expected and actual cursor locations. In at least one embodiment, a determination can be made that a player is using a secondary process or input because player input cursor 136 is not in a location in this game scene that would correspond to this determined action. In at least one embodiment, this position of this cursor 136 would then be considered an anomaly, as this position would not have caused this action to have occurred …).”
Regarding Claim 5. The combination of Pardeshi-Shepherd discloses the method according to 1, Pardeshi further discloses, “wherein the user perspective machine learning model is an artificial neural network (Pardeshi, Para [0001]: … At least one embodiment pertains to processing resources used to perform and facilitate artificial intelligence. For example, at least one embodiment pertains to processors or computing systems used to train neural networks according to various novel techniques described herein …).”
Regarding Claim 6. The combination of Pardeshi-Shepherd discloses the method according to claim 5, Shepherd further discloses, “wherein the step of determining the behavior pattern label and the indication to the likelihood of an anomaly occurring further comprises:
conducting data fusion based on the output of the user perspective machine learning model and the output of the behavior machine learning model (Shepherd, Para [0038]: … The trained machine learning model(s) 216 may represent a single model or an ensemble of base-level machine learning models, and may be implemented as any type of machine learning model 216. For example, suitable machine learning models 216 for use with the techniques and systems described herein include, without limitation, neural networks, tree-based models, support vector machines (SVMs), kernel methods, random forests, splines (e.g., multivariate adaptive regression splines), hidden Markov model (HMMs), Kalman filters (or enhanced Kalman filters), Bayesian networks (or Bayesian belief networks), expectation maximization, genetic algorithms, linear regression algorithms, nonlinear regression algorithms, logistic regression-based classification models, or an ensemble thereof. An “ensemble” can comprise a collection of machine learning models 216 whose outputs (predictions) are combined, such as by using weighted averaging or voting. The individual machine learning models of an ensemble can differ in their expertise, and the ensemble can operate as a committee of individual machine learning models that is collectively “smarter” than any individual machine learning model of the ensemble …).”
The motivation to further combine Shepherd remains same as in claim 1.
Regarding Claim 7. The combination of Pardeshi-Shepherd discloses the method according to claim 6, Shepherd further discloses, “wherein the data fusion step correlates the based on the output of the user perspective machine learning model and the output of the behavior machine learning model with respect to time (Shepherd, Para [0039]: … The training data that is used to train the machine learning model 216 may include various types of data 114. In general, training data for machine learning can include two components: features and labels… Example features included in the training data may include, without limitation, an amount of time a player spent playing video games 110 in general, an amount of time a player spent playing a particular video game 110, times of the day the player was logged in and playing video games 110 …); and
wherein the data fusion step comprises providing the output of the user perspective machine learning model and the output of the behavior machine learning model as inputs to a data fusion machine learning model (Shepherd, Para [0038]: … The trained machine learning model(s) 216 may represent a single model or an ensemble of base-level machine learning models, and may be implemented as any type of machine learning model 216 … An “ensemble” can comprise a collection of machine learning models 216 whose outputs (predictions) are combined, such as by using weighted averaging or voting. The individual machine learning models of an ensemble can differ in their expertise, and the ensemble can operate as a committee of individual machine learning models that is collectively “smarter” than any individual machine learning model of the ensemble …).”
Therefore it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to further combine the teachings of Shepherd into the teachings of Pardeshi, because it discloses that, “The individual machine learning models of an ensemble can differ in their expertise, and the ensemble can operate as a committee of individual machine learning models that is collectively “smarter” than any individual machine learning model of the ensemble (Shepherd, Para [0038])”.
Regarding Claim 8. The combination of Pardeshi-Shepherd discloses the method according to claim 1, Pardeshi further discloses, “wherein the step of determining a behavior pattern label and an indication to the likelihood of an anomaly occurring based on the received behavior data comprises the use of a prediction machine learning model (Pardeshi, Para [0092]: … In at least one embodiment, data center 700 may include tools, services, software or other resources to train one or more machine learning models or predict or infer information using one or more machine learning models according to one or more embodiments described herein. For example, in at least one embodiment, a machine learning model may be trained by calculating weight parameters according to a neural network architecture using software and computing resources described above with respect to data center 700. In at least one embodiment, trained machine learning models corresponding to one or more neural networks may be used to infer or predict information using resources described above with respect to data center 700 by using weight parameters calculated through one or more training techniques described herein …)”.
Regarding Claim 9. The combination of Pardeshi-Shepherd discloses the method according to claim 8, Pardeshi further discloses, “wherein the prediction machine learning model is an artificial neural network (Pardeshi, Para [0001, 0094]: … At least one embodiment pertains to processing resources used to perform and facilitate artificial intelligence … Inference and/or training logic 615 are used to perform inferencing and/or training operations associated with one or more embodiments. Details regarding inference and/or training logic 615 are provided below in conjunction with FIGS. 6A and/or 6B. In at least one embodiment, inference and/or training logic 615 may be used in system FIG. 7 for inferencing or predicting operations based, at least in part, on weight parameters calculated using neural network training operations, neural network functions and/or architectures, or neural network use cases described herein …), or a multi-headed hierarchical prediction model.
Regarding Claim 10. The combination of Pardeshi-Shepherd discloses the method according to claim 2, Pardeshi further discloses, “further comprising the step of:
receiving scene audio data, wherein the scene audio data is indicative of audio being played in a scene being presented by the electronic device (Pardeshi, Para [0058]: … In at least one embodiment, video segments may also be supplied to a labeling module 260. In at least one embodiment, this may be performed for each video segment processed by reconstruction module 256, or may be performed for segments in which a low reconstruction probability is assigned. In at least one embodiment, labeling module 260 processes segmented video clips using a machine learning-based technique to generate labels pertaining to scene, object, action, and audio recognition, referred to herein as a SOAR technique …), and wherein the step of determining the behavior pattern label comprises and the indication to the likelihood of an anomaly occurring further comprises:
processing the scene audio data using a scene audio machine learning model, and wherein the behavior pattern label and the indication to the likelihood of an anomaly occurring are additionally based on the output of scene audio artificial neural network (Pardeshi, Para [0064]: … In at least one embodiment, biometric data and other available data can be provided as input to a decision model to help improve a determination of cheating. In at least one embodiment, this can include camera data, audio data captured by a microphone, and any available biometric data such as heartbeat or motion. In at least one embodiment, a player looking away from a game for a period of time when player input is received may be indicative of cheating …).”
Regarding Claim 11. The combination of Pardeshi-Shepherd discloses the method according to claim 10, Pardeshi further discloses, “where the received scene audio data is processed to produce a time-frequency representation of the audio (Pardeshi, Para [0064]: … In at least one embodiment, biometric data and other available data can be provided as input to a decision model to help improve a determination of cheating. In at least one embodiment, this can include camera data, audio data captured by a microphone, and any available biometric data such as heartbeat or motion. In at least one embodiment, a player looking away from a game for a period of time when player input is received may be indicative of cheating. In at least one embodiment, changes in breathing patterns and heartrate may also be indicative of cheating. In at least one embodiment, a player may also make statements or utterances that are associated with cheating. In at least one embodiment, a speed or rate of input may also be analyzed, such as where a speed of input is above human capacity, or at least above a demonstrated capacity of a specific player. In at least one embodiment, input well above normal input rate for a player may also be indicative of potential cheating, such as through use of a bot or script …).”
Regarding Claim 12. The combination of Pardeshi-Shepherd discloses The method according claim 10, Pardeshi further discloses, “wherein the data fusion step is additionally based on the output of the scene audio machine learning model (Pardeshi, Para [0058]: … In at least one embodiment, video segments may also be supplied to a labeling module 260. In at least one embodiment, this may be performed for each video segment processed by reconstruction module 256, or may be performed for segments in which a low reconstruction probability is assigned. In at least one embodiment, labeling module 260 processes segmented video clips using a machine learning-based technique to generate labels pertaining to scene, object, action, and audio recognition …).”
Regarding Claim 13. The combination of Pardeshi-Shepherd discloses the method according to claim 1, Pardeshi further discloses, “further comprising the step of:
storing the behavior data as training data (Pardeshi, Para [0069, 0075], FIG. 6A-B: … In at least one embodiment, a data buffer may be used to house video and audio data for a game for a minimum period of time. In at least one embodiment, a hook or callback may be used, such as from an auto-highlights algorithm, to trigger cheating detection for any significant event or occurrence in a game. In at least one embodiment, a buffer can store at least a certain length of video data before and after an occurrence, such as 15 to 30 seconds before and after, in order to have sufficient data to analyze for cheating detection …);
updating the artificial neural networks based on the stored training data (Pardeshi, Para [0062]: … these clients can analyze gameplay during or post streaming, and can communicate results to a cheat server or other such target recipient. In at least one embodiment, these deployed weights can be updated over time as models get updated, such as through a transaction with a cheat server or with client updates …),
transmitting data indicative of the updated artificial neural networks to a centralized training server (Pardeshi, Para [0062]: … these clients can analyze gameplay during or post streaming, and can communicate results to a cheat server or other such target recipient. In at least one embodiment, these deployed weights can be updated over time as models get updated, such as through a transaction with a cheat server or with client updates …), and
receiving data indicative of a global machine learning model to update the artificial neural networks with (Pardeshi, Para [0357]: … In at least one embodiment, training pipeline 3204 (FIG. 32) may include a scenario where facility 3102 is training their own machine learning model, or has an existing machine learning model that needs to be optimized or updated. In at least one embodiment, imaging data 3108 generated by imaging device(s), sequencing devices, and/or other device types may be received. In at least one embodiment, once imaging data 3108 is received, AI-assisted annotation 3110 may be used to aid in generating annotations corresponding to imaging data 3108 to be used as ground truth data for a machine learning model …).”
Regarding Claim 14. The combination of Pardeshi-Shepherd discloses The method according to claim 4, Pardeshi further discloses, “further comprising the step of:
storing the user perspective data as training data (Pardeshi, Para [0069]: … In at least one embodiment, a data buffer may be used to house video and audio data for a game for a minimum period of time. In at least one embodiment, a hook or callback may be used, such as from an auto-highlights algorithm, to trigger cheating detection for any significant event or occurrence in a game. In at least one embodiment, a buffer can store at least a certain length of video data before and after an occurrence, such as 15 to 30 seconds before and after, in order to have sufficient data to analyze for cheating detection …);
updating the artificial neural networks based on the stored training data (Pardeshi, Para [0062]: … these clients can analyze gameplay during or post streaming, and can communicate results to a cheat server or other such target recipient. In at least one embodiment, these deployed weights can be updated over time as models get updated, such as through a transaction with a cheat server or with client updates …),
transmitting data indicative of the updated artificial neural networks to a centralized training server (Pardeshi, Para [0062]: … these clients can analyze gameplay during or post streaming, and can communicate results to a cheat server or other such target recipient. In at least one embodiment, these deployed weights can be updated over time as models get updated, such as through a transaction with a cheat server or with client updates …), and
receiving data indicative of a global machine learning model to update the artificial neural networks with (Pardeshi, Para [0357]: … In at least one embodiment, training pipeline 3204 (FIG. 32) may include a scenario where facility 3102 is training their own machine learning model, or has an existing machine learning model that needs to be optimized or updated. In at least one embodiment, imaging data 3108 generated by imaging device(s), sequencing devices, and/or other device types may be received. In at least one embodiment, once imaging data 3108 is received, AI-assisted annotation 3110 may be used to aid in generating annotations corresponding to imaging data 3108 to be used as ground truth data for a machine learning model …).”
Regarding Claim 15. The combination of Pardeshi-Shepherd discloses The method according to claim 10, Pardeshi further discloses, “further comprising the step of:
storing the scene audio data as training data (Pardeshi, Para [0069]: … In at least one embodiment, a data buffer may be used to house video and audio data for a game for a minimum period of time. In at least one embodiment, a hook or callback may be used, such as from an auto-highlights algorithm, to trigger cheating detection for any significant event or occurrence in a game. In at least one embodiment, a buffer can store at least a certain length of video data before and after an occurrence, such as 15 to 30 seconds before and after, in order to have sufficient data to analyze for cheating detection …);
updating the artificial neural networks based on the stored training data (Pardeshi, Para [0062]: … these clients can analyze gameplay during or post streaming, and can communicate results to a cheat server or other such target recipient. In at least one embodiment, these deployed weights can be updated over time as models get updated, such as through a transaction with a cheat server or with client updates …),
transmitting data indicative of the updated artificial neural networks to a centralized training server (Pardeshi, Para [0062]: … these clients can analyze gameplay during or post streaming, and can communicate results to a cheat server or other such target recipient. In at least one embodiment, these deployed weights can be updated over time as models get updated, such as through a transaction with a cheat server or with client updates …), and
receiving data indicative of a global machine learning model to update the artificial neural networks with (Pardeshi, Para [0357]: … In at least one embodiment, training pipeline 3204 (FIG. 32) may include a scenario where facility 3102 is training their own machine learning model, or has an existing machine learning model that needs to be optimized or updated. In at least one embodiment, imaging data 3108 generated by imaging device(s), sequencing devices, and/or other device types may be received. In at least one embodiment, once imaging data 3108 is received, AI-assisted annotation 3110 may be used to aid in generating annotations corresponding to imaging data 3108 to be used as ground truth data for a machine learning model …).”
Regarding Claim 16. The combination of Pardeshi-Shepherd discloses The method according to claim 1, Pardeshi further discloses, “further comprising the step of:
receiving peripheral data, wherein the peripheral data is data indicative of the current state of a peripheral that is coupled to the electronic device (Pardeshi, Para [0263-00264]: … In at least one embodiment display device 2311 can include one or more of an internal display device, as in a mobile electronic device or a laptop device or an external display device attached via a display interface (e.g., DisplayPort, etc.) …).”
Regarding Claim 18. The combination of Pardeshi-Shepherd discloses the method according to claim 1, Pardeshi further discloses, “further comprising the step of:
recognizing unexpected patterns based on the received behavior data, wherein unexpected patterns include at least one or more of the following: too many inputs being provided at a given time, to location of the inputs being provided at a given time are too distal for a hand of the user, and the speed of inputs being too fast for the hand of the user (Pardeshi, Para [0050-0051]: … In at least one embodiment, a position of a cursor 136 used for player input does not match an expected position of that cursor to provide input that matches this swing of golf club 134. In at least one embodiment, this can be measured as a delta between expected and actual cursor locations. In at least one embodiment, a determination can be made that a player is using a secondary process or input because player input cursor 136 is not in a location in this game scene that would correspond to this determined action. In at least one embodiment, this position of this cursor 136 would then be considered an anomaly, as this position would not have caused this action to have occurred … In at least one embodiment, there may be other information available that may be indicative of cheating. In at least one embodiment, a player may have a webcam (or other camera) available that captures at least a portion of that player during gameplay. In at least one embodiment, that video data can be analyzed and correlated with gameplay. In at least one embodiment, a video 162 of a player may show that player looking away from a game display for a period of time during which multiple inputs are provided that are highly unlikely to be performed by a player not paying attention to that game. In at least one embodiment, this may include a care making a precise turn into a lane as indicated in video frame 160 of FIG. 1C. In at least one embodiment, a similar determination can be made if no player input is detected in corresponding game video, such as no moving of a cursor or other such input…).”
Regarding Claim 19. The combination of Pardeshi-Shepherd discloses the method according to claim 1, Pardeshi further discloses, “further comprising the step of:identifying an unexpected pattern based on the received behavior data, wherein the unexpected pattern indicates a sequence of the users inputs are beyond a physical limitation of the user (Pardeshi, Para [0066]: … In at least one embodiment, identified objects can also include those corresponding to user input, such as may include use of a cursor 308. In at least one embodiment, these labels can help give context to an identified anomaly for use in determining whether that anomaly likely corresponds to cheating. In at least one embodiment, a reconstruction module finding a confident anomaly may have keywords associated from a labeling module that give context to that anomaly, such as a cursor in a location away from a normal location for a swing of a golf club generating a specific motion of a golf ball in a scene of golf In at least one embodiment, this may be indicative of use of an auto-combo tool. In at least one embodiment, if this segment comes from a highlight video and corresponds to a hole in one, that may further be used with this labeling to make a determination that a player is using some type of improper assistance to play this game and gain an unfair advantage …)”.
Regarding Claim 20. The combination of Pardeshi-Shepherd discloses the method according to claim 19, Pardeshi further discloses, “wherein the physical limitations include any one or more of the following:
too many button presses at a given time,
a series of button presses is too quick, and
button presses which are physically impossible for a user (Pardeshi, Para [0050]: … In at least one embodiment, a position of a cursor 136 used for player input does not match an expected position of that cursor to provide input that matches this swing of golf club 134. In at least one embodiment, this can be measured as a delta between expected and actual cursor locations. In at least one embodiment, a determination can be made that a player is using a secondary process or input because player input cursor 136 is not in a location in this game scene that would correspond to this determined action. In at least one embodiment, this position of this cursor 136 would then be considered an anomaly, as this position would not have caused this action to have occurred …)”.
Regarding Claim 21. The combination of Pardeshi-Shepherd discloses the method according to claim 19, Pardeshi further discloses, “wherein the received behavior data comprises gyroscope data from a peripheral associated with the electronic device and wherein the step of identifying the unexpected pattern is based on the gyroscope data (Pardeshi, Para [0110]: … In at least one embodiment, other components may be communicatively coupled to processor 910 through components discussed above. In at least one embodiment, an accelerometer 941, Ambient Light Sensor (“ALS”) 942, compass 943, and a gyroscope 944 may be communicatively coupled to sensor hub 940 …).”
Regarding Claim 22. The combination of Pardeshi-Shepherd discloses the method according to claim 21, wherein the step of identifying the unexpected pattern comprises identifying a button press that physically could not have occurred based on the gyroscope data (Pardeshi, Para [0050, 0110]: … In at least one embodiment, a position of a cursor 136 used for player input does not match an expected position of that cursor to provide input that matches this swing of golf club 134. In at least one embodiment, this can be measured as a delta between expected and actual cursor locations. In at least one embodiment, a determination can be made that a player is using a secondary process or input because player input cursor 136 is not in a location in this game scene that would correspond to this determined action. In at least one embodiment, this position of this cursor 136 would then be considered an anomaly, as this position would not have caused this action to have occurred … In at least one embodiment, other components may be communicatively coupled to processor 910 through components discussed above. In at least one embodiment, an accelerometer 941, Ambient Light Sensor (“ALS”) 942, compass 943, and a gyroscope 944 may be communicatively coupled to sensor hub 940 …), and
wherein the button press that physically could not have occurred based on the gyroscope data is indicative of a counterfeit or third party input device (Pardeshi, Para [0050, 0110]: … )”.
Regarding Claim 23. The combination of Pardeshi-Shepherd discloses the method according to claim 1, Shepherd further discloses, “wherein the behavior pattern label and the behavior pattern label hierarchies relate to a skill of the user of the electronic device (Shepherd, Para [0036, 0043]: … In some embodiments, other factors (e.g., skill level, geographic region, etc.) are considered in the matchmaking process, which may cause further breakdowns and/or subdivisions of players into a fewer or greater number of matches 218 … As yet another example, the trained machine learning model(s) 216 may be configured to output a trust score 118 that relates to a probability of a player behaving, or not behaving, in accordance with a “high-skill” behavior. In this manner, the scoring can be used to identify highly-skilled players, or novice players, from a set of players. This may be useful to prevent situations where experienced gamers create new user accounts pretending to be a player of a novice skill level just so that they can play with amateur players …).”
The motivation to further combine Shepherd remains same as in claim 1.
Regarding Claim 24. The combination of Pardeshi-Shepherd discloses the method according to 23, Shepherd further discloses, “wherein the behavior pattern label hierarchies comprise beginner, intermediate, and expert (Shepherd: Para [0036, 0043]”.
The motivation to further combine Shepherd remains same as in claim 23.
Regarding Claim 25. The combination of Pardeshi-Shepherd discloses the method according to claim 1, Pardeshi further discloses, “wherein an anomaly is indicative of the user cheating at a game being played on the electronic device (Pardeshi, Para [0048]: … a user may access digital content through a computing device. In at least one embodiment, this digital content may include gaming content for one or more games. In at least one embodiment, these games may be played in an online manner, in which one or more players access game content over one or more networks served by one or more game servers. In at least one embodiment, these players may compete against one another, with a potential for rewards, compensation, or even just bragging rights. In at least one embodiment, these benefits may cause certain players to attempt to cheat or otherwise gain unfair advantage in these games … ).”
Regarding Claim 27. This claim contains all the same or similar limitations as claim 1, and hence similarly rejected as claim 1.
*** Note: Pardeshi also discloses non-transitory computer-readable storage medium comprising instructions executed by a computer (Pardeshi: Para [0421]).
Regarding Claim 28. This claim contains all the same or similar limitations as claim 1, and hence similarly rejected as claim 1.
*** Note: Pardeshi also discloses a system (Pardeshi: Para [0052], FIG. 2A).
Claim 17 is rejected under 35 U.S.C. 103 as being unpatentable over Pub. No.: US 2021/0232907 A1 to Pardeshi et al. (hereinafter “Pardeshi”) in view of Pub. No.: US 2021/0038979 A1 to Bleasdale-Shepherd et al. (hereinafter “Shepherd”), as applied to claim 16 above, and further in view of Pub. No.: US 20140274242 A1 to HASWELL (hereinafter “HASWELL”).
Regarding Claim 17. The combination of Pardeshi-Shepherd discloses The method according to claim 16, Pardeshi further discloses, “wherein the peripheral data comprises an indication of a type of the peripheral and an identifier of the peripheral (Pardeshi, Para [0108, 0263]: … In at least one embodiment display device 2311 can include one or more of an internal display device, as in a mobile electronic device or a laptop device or an external display device attached via a display interface (e.g., DisplayPort, etc.). In at least one embodiment, display device 2311 can include a head mounted display (HMD) such as a stereoscopic display device for use in virtual reality (VR) applications or augmented reality (AR) applications …); and
However, the combination of Pardeshi-Shepherd does not explicitly teach, but HASWELL from same or similar field of endeavor teaches:
“wherein the received peripheral data is compared with a list of known peripheral data, and wherein the detection of the behavior pattern label and the indication to the likelihood of an anomaly occurring is further based on this comparison (HASWELL, FIG. 4, Abstract, Para [0016]: … According to embodiments of the invention, player skill may be metered at distance or time increments having any chosen duration, however it is preferred to meter player skill in digital environments at the highest possible fidelity. This degree of unique data visibility facilitates the following functions, in a modular system configuration: Accurate, real-time, player skill quantification; 2. Accurate player skill matching with minimal player history; 3. Optimization of player engagement through real time experience curves; 4. Player liquidity management; 5. Compliance monitoring for both operator and competent authority; and 6. Detection of cheating and fraudulent player behavior patterns … Apparatus and methods are described for operating an online interactive simulation/game environment, where an object's motion is controlled by a player along a path. A player's skill level is quickly quantified as they control the object traversing the path, so that the player can be properly placed in games with those of like skill, even prior to having competed with others. An optimal path is established for the object's travel during a portion of a game, and at each increment of travel, an optimum velocity or time delta is established. The path of the object being controlled by a player being rated is then tracked over the same path. At each distance increment its position and velocity or time delta are recorded and compared with the optimum. Deviations therebetween are calculated on an incremental basis, and the aggregate determines the player's skill level for a set of equivalent conditions …; Examiner’s Interpretation: Fig. 4 shows 3 levels of shills - "beginners" vs. "intermediates" vs. "experts" and the peripheral data inputs are steering, throttle, brake and so on as in FIG. 4).”
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 teachings of HASWELL into the combined teachings of Pardeshi-Shepherd, because it discloses that, “When first placing a player in a multi-player competition, it is critical for encouraging and maintaining player participation that their initial experience is as positive as possible. Therefore, it is advantageous that a player be matched against other players of similar skill, even in their very first competition. Therefore their skill must be known prior to their first competition. Then during the player's first competition they will enjoy a fair and competitive matching with other players, and simultaneously will be re-evaluated on their skill as the game progresses so that they will subsequently be optimally matched for their second, and subsequent competitions (SHARIFF, Para [0085])”.
Claim 29 is rejected under 35 U.S.C. 103 as being unpatentable over Pub. No.: US 2021/0232907 A1 to Pardeshi et al. (hereinafter “Pardeshi”) in view of Pub. No.: US 2021/0038979 A1 to Bleasdale-Shepherd et al. (hereinafter “Shepherd”), as applied to claim 28 above, and further in view of Pub. No.: US 20230251950 A1 to SHARIFF et al. (hereinafter “SHARIFF”).
Regarding Claim 29. The combination of Pardeshi-Shepherd discloses the system according to claim 28, Pardeshi further discloses, “wherein the system further comprises a training server configured to receive a model update from an electronic device, wherein the training server is configured to receive multiple model updates from further electronic devices (Pardeshi, Para [0042, 0047, 0062, 0369, 0398-0399, 0407-0408], fig. 35: … FIG. 32 is a system diagram for an example system for training, adapting, instantiating and deploying machine learning models in an advanced computing pipeline, in accordance with at least one embodiment … In at least one embodiment, decision data from decision module 258 can be provided as input to a cheat server 262. In at least one embodiment, cheat server 262 can maintain a profile of player history with data about decisions about cheating, including false positives, by gaming session … In at least one embodiment, these clients can analyze gameplay during or post streaming, and can communicate results to a cheat server or other such target recipient. In at least one embodiment, these deployed weights can be updated over time as models get updated, such as through a transaction with a cheat server or with client updates. In at least one embodiment, a cost of data acquisition is minimized by using gameplay videos of players that use such a platform for analysis and training, with permission being obtained as appropriate … In at least one embodiment, system 3200 (e.g., training system 3104 and/or deployment system 3106) may implemented in a cloud computing environment (e.g., using cloud 3226) …); and
wherein the training server is configured to conduct [federated] learning using at least the model update to generate a [global] model, and provide the [global] model to the electronic device (Pardeshi, Para [0407-0409], FIG. 35: … FIG. 35A illustrate a data flow diagram for a process 3500 to train, retrain, or update a machine learning model … model training 3114 may include retraining or updating an initial model 3504 (e.g., a pre-trained model) using new training data (e.g., new input data, such as customer dataset 3506, and/or new ground truth data associated with input data). In at least one embodiment, to retrain, or update, initial model 3504, output or loss layer(s) of initial model 3504 may be reset, or deleted, and/or replaced with an updated or new output or loss layer(s). In at least one embodiment, initial model 3504 may have previously fine-tuned parameters (e.g., weights and/or biases) that remain from prior training, so training or retraining 3114 may not take as long or require as much processing as training a model from scratch. In at least one embodiment, during model training 3114, by having reset or replaced output or loss layer(s) of initial model 3504, parameters may be updated and re-tuned for a new data set based on loss calculations associated with accuracy of output or loss layer(s) at generating predictions on new, customer dataset 3506 (e.g., image data 3108 of FIG. 31) …).”
However, the combination of Pardeshi-Shepherd does not explicitly teach, but SHARIFF from same or similar field of endeavor teaches, “conduct federated learning using at least the model update to generate a global model, and provide the global model to the electronic device (SHARIFF, Para [0066, 0091]: … FIG. 19 is a diagram illustrating an example process of Federated Learning (FL), according to various embodiments. In the federated learning setup, all devices for example the electronic devices (120A-120N) and the smart devices (130A-130N) report the trained model to the server 110 for aggregation, wherein the aggregated global model is shared back to all devices for identifying new behaviours. A behavior classification model is downloaded from the server to selected mobile devices, wherein the model is trained using the user’s personal data on the device. Weight-updates, after training locally, are communicated to the server. Aggregators average out the learnings from various devices and update the final model in server side …)”
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 teachings of SHARIFF into the combined teachings of Pardeshi-Shepherd, because it discloses that, “Various example embodiments disclose that the method further includes: identifying, by the server, outlier and anomalous behaviors which have probabilities different from global averages (SHARIFF, Para [0015])”.
Pertinent Prior Arts
The following prior arts made of record and not relied upon are considered pertinent to applicant's disclosure.
US 20210308586 A1; Mulasmajic et al.: Mulasmajic discloses systems, methods, and computer storage media directed to anti-cheat detection in online multiplayer video games. An anti-cheat kernel driver adapted to validate and secure a system state of a gaming device is loaded during a booting process of the gaming device. The loaded anti-cheat kernel driver ensures that the system state cannot be tampered with from the time of boot through the duration of gameplay. The loaded anti-cheat kernel driver can also receive anti-cheat modules communicated to the gaming device from an anti-cheat server, so that anti-cheat modules are received and executed on an ad hoc basis dictated by the anti-cheat server. The anti-cheat server can autonomously manage the anti-cheat operations performed on a kernel layer or application layer of the gaming device, and conduct anti-cheat mitigation operations if necessary.
US 11980820 B2; Tsuria: Tsuria discloses Techniques for countering cheats in a multi-player gaming environment are described herein. In accordance with various embodiments, a server includes a cheating analyzer running on processor(s) and a non-transitory memory for storing cheat data and counter cheat data. The server identifies suspected cheating in a game and determines a probability of the suspected cheating based on the cheat data. The server also selects a counter cheat of the suspected cheating corresponding to the probability value based on the cheat data and the counter cheat data. The server then applies the counter cheat approximate the probability value in the game. In some embodiments, the server deploys local copies of the cheating analyzer to client devices, including sending at least a portion of the cheat data and/or counter cheat data. The local copies facilitate the cheat identification, the probability determination, the counter cheat selection, and/or the application of the counter cheat.
US 20180361250 A1; Brew et al.: Brew discloses a computer-implemented method for monitoring game activity of a game system. A non-limiting example of the method includes monitoring, by a processing device, game activity of the game system. The processing device determines characteristics of the game activity, along with expected characteristics of the game activity. The processing device analyzes the characteristics of the game activity and the expected characteristics of the game activity. Based at least in part on analyzing the characteristics of the game activity and the expected characteristics of the game activity, an entity that is controlling the game system is determined.
The present invention generally relates to electronic games, and more specifically, to monitoring electronic game activity to detect that an electronic game is being controlled or played by a surrogate computer system (e.g., a game bot).
US 20220180173 A1; Jonnalagadda et al.: Jonnalagadda discloses Apparatuses, systems, and techniques to detect cheating in a computer game. In at least one embodiment, one or more circuits use one or more neural networks to detect cheating by one or more users of a computer game based, at least in part, on one or more images generated by the computer game.
At least one embodiment pertains to graphics processing units configured to protect against use of illicit information in images using neural networks. For example, at least one embodiment pertains to operations encountered in training and using neural networks, executed on Graphics Processing Units, to protect against use of cheating or other illicit information within images displayed to users of computing devices.
US 20230241512 A1; Kim et al.: Kim discloses a cheating detection strategy for interactive programs, which detects programmatically-generated motion from actual human-generated motion based on a comparison of actual motion data to inferred motion data. The cheating detection strategy uses visual and input information to ensure that the input matches the output to detect and avoid cheating tools positioned in between the input and the output. In one example, the disclosure provide a method of monitoring cheating in interactive programs that includes: (1) receiving actual motion data from a user input device, wherein the actual motion data corresponds to interacting with the interactive program, (2) receiving image data of the interactive program that includes image sequences of the interactive program to display on a screen, (3) comparing the actual motion data to inferred motion data determined from the image sequences, and (4) determining possible cheating based on the comparing.
This application is directed, in general, to detecting cheating in video game play and, more specifically, to detecting cheating at a monitor that displays the video game.
US 20180182208 A1; LIU et al.: LIU discloses systems and methods for detecting cheating at a game platform level using machine learning techniques. One example provides a computing system comprising a logic subsystem and a data-holding subsystem. The data-holding subsystem comprises instructions executable by the logic subsystem to receive notifications related to user progress in a game provided by the game to the online game platform, apply a classifying function to classify the user progress in the game as normal or outlying based upon the notifications, if the progress is classified as outlying then taking an action in response to the outlying classification, and if the progress is not classified as outlying then not taking the action.
US 20220001283 A1; Lundquist et al.: Lundquist discloses An anti-cheat system that may be accessed over a network and stored directly into volatile memory of a user computing system. In some embodiments, this anti-cheat system may scan, or access portions of, the volatile memory of the user computing system to detect whether cheat software or other unauthorized software that may interact with a game application is detected on the user computing system. The accessed portions of the volatile memory may be compared with one or more signatures that are associated with the execution of cheat software on a computing system. The anti-cheat system may be prevented from being stored within non-volatile memory, thereby preventing malicious users from modifying the anti-cheat system.
Conclusion
Applicant's amendment necessitated the new ground(s) of rejection presented in this Office action. Accordingly, THIS ACTION IS MADE FINAL. See MPEP § 706.07(a). Applicant is reminded of the extension of time policy as set forth in 37 CFR 1.136(a).
A shortened statutory period for reply to this final action is set to expire THREE MONTHS from the mailing date of this action. In the event a first reply is filed within TWO MONTHS of the mailing date of this final action and the advisory action is not mailed until after the end of the THREE-MONTH shortened statutory period, then the shortened statutory period will expire on the date the advisory action is mailed, and any nonprovisional extension fee (37 CFR 1.17(a)) pursuant to 37 CFR 1.136(a) will be calculated from the mailing date of the advisory action. In no event, however, will the statutory period for reply expire later than SIX MONTHS from the mailing date of this final action.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to MAHABUB S AHMED whose telephone number is (571)272-0364. The examiner can normally be reached on 9AM-5PM EST M-F.
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, Kambiz Zand can be reached on (571)272-3811. The fax phone number for the organization where this application or proceeding is assigned is 571-273-8300.
Information regarding the status of an application may be obtained from the Patent Application Information Retrieval (PAIR) system. Status information for published applications may be obtained from either Private PAIR or Public PAIR. Status information for unpublished applications is available through Private PAIR only. For more information about the PAIR system, see http://pair-direct.uspto.gov. Should you have questions on access to the Private PAIR system, contact the Electronic Business Center (EBC) at 866-217-9197 (toll-free). If you would like assistance from a USPTO Customer Service Representative or access to the automated information system, call 800-786-9199 (IN USA OR CANADA) or 571-272-1000.
/MAHABUB S AHMED/Examiner, Art Unit 2434
/KAMBIZ ZAND/Supervisory Patent Examiner, Art Unit 2434
Prosecution Insights