DETAILED ACTION
A. This action is in response to the following communications: Amendment filed: 03/03/2026. This action is made Final.
B. Claims 1-20 remain pending.
Claim Rejections - 35 USC § 102
In the event the determination of the status of the application as subject to AIA 35 U.S.C. 102 and 103 (or as subject to pre-AIA 35 U.S.C. 102 and 103) is incorrect, any correction of the statutory basis (i.e., changing from AIA to pre-AIA ) for the rejection will not be considered a new ground of rejection if the prior art relied upon, and the rationale supporting the rejection, would be the same under either status.
The following is a quotation of the appropriate paragraphs of 35 U.S.C. 102 that form the basis for the rejections under this section made in this Office action:
A person shall be entitled to a patent unless –
(a)(1) the claimed invention was patented, described in a printed publication, or in public use, on sale, or otherwise available to the public before the effective filing date of the claimed invention.
Claim(s) 1,3-8,10-15 and 17-20 is/are rejected under 35 U.S.C. 102(a)(1) as being clearly anticipated by Le Chevalier (US Pub. 2024/0281052 A1), herein referred to as “Le Chevalier”.
As for claims 1, 8 and 15, Le Chevalier teaches. A method and corresponding system of claim 8 and computer product of claim 15, implemented by a processor coupled to a memory; specifically to claim 8 a set of instructions stored in memory and executed by at least one of the processors to perform actions; specifically for claim 15 a computer readable storage medium comprising a second of instructions that, when executed by a processor, are effective to perform the actions, comprising (fig. 1; par. 25-26 describe the hardware environment configurations that implement the invention):
displaying a plurality of avatars locating in a first virtual area that is provided by a virtual reality (VR) platform, wherein the avatars are visible to a first plurality of users with avatars in the first virtual area, and wherein each of the users controls one of the displayed avatars (par. 16 users participate in virtual and augmented reality environments/worlds which as a collective is termed metaverse wherein a user represents their persona through an avatar also referred to as “online avatar”; wherein each avatar can be personalized, life-like and/or animated. Through avatars, users can collectively and simultaneously participate in specific events, be present in particular virtual places, and participate in activities with other users with individual agency. Also note that avatars can replicate an entire user and/or their body movements to create a vivid feeling of the user being physically present in the metaverse);
receiving a request from a selected user from the plurality of users, wherein the selected user controls a first avatar from the plurality of avatars (par. 17 the primary avatar can be a live virtual representation of the associated user such that the actions of the primary avatar are defined and controlled in real-time by the user (e.g., via the user's engagement with a user input such as a device controller, touch screen, and/or keyboard);
in response to the request: cloning the first avatar creating a second avatar, wherein the second avatar is visible to the selected user and invisible to one or more of the plurality of users, and wherein the selected user controls the second avatar; and (par. 18 a user can clone their primary avatar (virtual self), herein clone, wherein the clone can be in any social space, domain and/or event separate from the primary avatar (virtual self) that the user is interactively controlling; par. 19 the user’s virtual self can be cloned multiple times and is saved as a set of avatars, this set of clones can be sent to multiple virtual worlds (first, second to the nth virtual area), the user can participate interactively in any of the virtual worlds where a current clone is visiting from the set of clones that are actively placed in; thus the user’s primary avatar is only visible in one virtual world at a time while the clone avatar is visiting/representing the primary avatar in multiple virtual worlds simultaneously, thus when the user wishes to leave they can move from one world where other users that were actively engaging with the user and their primary avatar (virtual self) and hence “visible” (real person interacts with real person) to these other users, the primary avatar can transport to another virtual world yet leaving behind a clone and now becoming “invisible” to the other users in the first virtual area because the real user of the primary avatar is no longer in the first virtual world only the clone that represents them as the claim also suggests. The user no longer will be interacting with users in the first area only their clone will be present in the first area, wherein described further in Le Chevalier clones can be operated in different modes passive/interactive; the reason for the user leaving is moot but the functionality remains the same), wherein a trained voice model of the selected user is trained on voice samples of the selected user (paragraph 46 can be configured to analyze the audio including the voice communication, determine a reply (e.g., based on assigned characteristics or rules), and present the reply (e.g., by delivering pre-recorded or dynamic content generated through a text-to-speech engine), similar to the operation of a bot conversation. The mention of assigned characteristics or rules implies from paragraph 48 that recordings of the user are used to train an artificial intelligence model of how a user vocals output in order to clone the user);
“…As described above, the memory 202 can store software code representing instructions for the cloned avatar modeling and personalization module 213. In some implementations, the processor 201 can execute the instructions of the cloned avatar modeling and personalization module 213 to manage the generation and/or storing of characteristics of the cloned avatars that are related to the look and feel of each cloned avatar. Cloned avatars of a user can be modeled based on a primary avatar of a user and can optionally be personalized by the user (e.g., based on their respective passive or active roles and/or expected activities)…”
automatically controlling the first avatar using artificial intelligence (AI), wherein the selected user no longer controls the first avatar (par. 41 the cloned avatar roles and properties module can assign (e.g., under the control of a user and/or using machine learning and/or analytics algorithms) a passive role or mode or an interactive role or mode to a cloned avatar; par. 42 describes further the interactive mode which was set by machine learning for automation of clone interaction within different virtual worlds for different purposes and different communication interactions between other avatars be them real/clones).
[0019] …Thus, a user of the metaverse may simultaneously virtually participate in and/or be represented in distinct worlds (e.g., distinct virtual reality applications) of a metaverse using the set of avatars. Additionally, using an avatar of the set of avatars, the user can virtually participate and/or be represented in a world (e.g., a virtual reality application) of a metaverse without needing to actively control the avatar through actively interacting in real-time with a user input (e.g., via physically interacting with a controller device) of a device running the virtual reality application in the real world. For example, the user can virtually participate and/or be represented in a world by a cloned avatar from the set of avatars when the user is offline and unable to actively and in real-time control the user's primary avatar. The user can also virtually participate and/or be represented in a world by a cloned avatar from the set of avatars when the user is actively controlling the user's primary avatar in real-time (e.g., for a different purpose and/or in a different world than the cloned avatar).
and emitting vocalization corresponding to the first avatar using the trained voice model having been trained on the voice samples of the selected user (In paragraph 49 further teaches using the model to emulate voice output from a cloned avatar (emulate one or more user inputs that an online user would engage with when controlling the primary avatar of the user, such as by interacting… for example speaking into a microphone). Further based upon the modeled cloned avatar the system is able to emulate in passive or active mode to provide automated voice replies; these replies were modeled off of user audio inputs (voice samples per se).
“…A passive or active cloned avatar can provide automated voice replies to queries through, for example, a text-to-speech API, such that the cloned avatar communicates with other avatars similarly to when an online user speaks into a microphone and the recorded speech is associated with an avatar of the user (e.g., in real time) to communicate with another avatar…”
Paragraph 50 describes the storage solution to the audio/video capture systems 215 wherein frames captured are deconstructed into images and audio and analyzed separately which described in paragraph 51 this information is passed to the AI navigation and control systems module 216 to simulate behavior of the user through movement and voice outputs. The rest of paragraph 51 describes training the model that represents the users cloned avatar by analyzing the audio/video).
“…For example, after the control interface systems module 217 emulates a physical interaction between the user and a user input (e.g., a physical input controller), the AI navigation and control systems module 216 can analyze the outcome of a rendering engine associated with the virtual reality application (e.g., via analyzing data captured by the audio/video capture systems module 215) to determine an effect, if any, on and/or by the avatar resulting from the emulated physical interaction, and determining what, if anything, the cloned avatar should do next. For example, the analysis and determination by the AI navigation and control systems module 216 can be based on real-time inputs from, for example, the audio/video capture systems module 215. The AI navigation and control systems module 216 can send an instruction to the controller interface systems module 217 based on the determination of what the cloned avatar should do next so that the controller interface systems module 217 can emulate a physical interaction between the user and a user input to cause the cloned avatar to take a particular action (e.g., move, respond to another avatar, etc.)…”
As for claims 3, 10 and 17, Le Chevalier teaches. The method of claim 1 and corresponding claims 8 and 15 further comprising:
moving the first avatar using a trained non-verbal model that is utilized by the AI (par. 51 use of an AI navigation for move clone avatar and how to interact with objects and other avatars as directed by the AI navigation through trained learning utilizing saved models), wherein the trained non-verbal model is trained by collecting non-vocal mannerisms and non-verbal communications while the selected user controls an avatar in the VR platform ( par. 56,58,62 and 63 the users movement is recorded and used in the training of the model such that the cloned avatar mimics the movement of the user; The plurality of characteristics of the object includes at least one of an identifier, a function, whether it stays in one location or moves, a size, a color, a current movement status, an estimated speed, and a location. In some implementations, the first machine learning model can be an object detection machine learning model. The machine learning model can include, for example, methods from one or more region-based convolutional neural networks (R-CNNs) (e.g., R-CNN, Fast R-CNN, and Faster-RCNN), methods from the You Only Look Once (YOLO) family of object recognition models (e.g., YOLOv2, YOLOv3), methods from the HydraNet network architecture template, and/or any other suitable methods or techniques.At 308, the example cloned avatar management process 300 includes predicting, at the processor, at least one movement parameter of each object, based on the data and the plurality of characteristics of that object. In some implementations, the at least one movement parameter of the object includes at least one of a moving direction, a speed, and a likelihood of movement in a pre-determine time period. The machine learning model can output a set of multicam features about the pixel that can be used for predictions. To account for time and/or movement, a transformer module of the machine learning model can generate multiscale features that are concatenated with kinematics, such as velocity and acceleration, and position tracking of the cloned avatar. The feature queue(s) can be consumed by a spatial recurrent neural network (RNN) that tracks the location of the cloned avatar and relevant objects, avatars, and/or indicators of possible cloned avatar actions and can selectively read and write to a memory or cached queue to generate a map as the cloned avatar moves about the environment.)
As for claims 4, 11 and 18, Le Chevalier teaches. The method of claim 1 and corresponding claims 8 and 15 further comprising:
after a period of time subsequent to the cloning, deleting the first avatar, wherein the first avatar is no longer visible to the first plurality of users (par. 112 duration of time is tracked for various activities by the clone in such that a start time and end time can be set for a clone which suggest the clone will not be present in a location after a period of time).
As for claims 5, 12 and 19, Le Chevalier teaches. The method of claim 1 and corresponding claims 8 and 15 further comprising:
moving, based on one or more actions received from the selected user (par. 41 user selects passive or interactive clone mode), the second avatar to a second virtual area provided by the VR platform, wherein the second virtual area is virtually distant from the common virtual area so that first plurality of users with avatars in the first virtual area cannot view the second avatar in the second virtual area; and visualizing the second avatar so that the second avatar is no longer invisible to a second plurality of users with avatars in the second virtual area (par. 19 The user can also virtually participate and/or be represented in a world by a cloned avatar from the set of avatars when the user is actively controlling the user's primary avatar in real-time (e.g., for a different purpose and/or in a different world than the cloned avatar); thus the user is able to not be visible to set of first users in a first area because that primary avatar and user has left first area to a second area and left behind a passive/interactive clone which can be removed after a duration of time set in the settings).
As for claims 6, 13 and 20, Le Chevalier teaches. The method of claim 5 and corresponding claims 12 and 19 further comprising:
receiving, at a microphone, a plurality of vocalizations from the selected user;
sending the received vocalizations to the first plurality of users prior to the cloning; inhibiting the sending of the received vocalizations to the first plurality of users after the cloning; and training an artificial intelligence voice model corresponding to the selected user using the received vocalizations (par. 30 audio sensors for user input; par. 38 items 211-217 which included captured audio/video to use AI modeling in creating clones of the primary avatar).
As for claims 7, 14 and 20, Le Chevalier teaches. The method of claim 6 and corresponding claims 13 and 20 further comprising:
after the visualizing of the second avatar, sending the received vocalizations to the second plurality of users (par. 43 audio communication via clone from AI training).
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) 2, 9 and 16 is/are rejected under 35 U.S.C. 103 as being unpatentable over Le Chevalier in view of Ma, Long et al. (US Pub. 2020/0043471 A1), herein referred to as “Ma”.
As for claims 2, 9 and 16, Le Chevalier teaches. The method of claim 1 and corresponding claims 8 and 15 wherein the first avatar emits the vocalization based on collecting in the VR platform the voice samples of the selected user speaking into a microphone (par. 50-51 collecting voice input/ samples of user to train AI model of user for creating clone avatar, all voice inputted is utilized)).
Le Chevalier does not specifically teach randomly collecting voice input; however in the same field of endeavor Ma teaches randomly collecting voice input (par. 92 randomly generating 10 groups of data (each group of data including 10 voice samples with identical content respectively provided by 10 unrepeated persons) as training sets, randomly selecting voice data of 7 persons from each set for training the model parameters (Eps, MinPts) to maximize the JC, using voice data of the remaining 3 persons for verification to alleviate model over-fitting).
It would have been obvious to one of ordinary skill in the art before the effective filing date to combine Le Chevalier into Ma because Ma suggest in paragraph 92 to randomly select voice data to alleviate model over-fitting.
(Note :) It is noted that any citation to specific, pages, columns, lines, or figures in the prior art references and any interpretation of the references should not be considered to be limiting in any way. A reference is relevant for all it contains and may be relied upon for all that it would have reasonably suggested to one having ordinary skill in the art. In re Heck, 699 F.2d 1331, 1332-33, 216 USPQ 1038, 1039 (Fed. Cir. 1983) (quoting In re Lemelson, 397 F.2d 1006,1009, 158 USPQ 275, 277 (CCPA 1968)).
Response to Arguments
Applicant's arguments filed 03/03/2026 have been fully considered but they are not persuasive.
A1. Applicant argue that the prior art does not teach wherein a trained voice model of the selected user is trained on voice samples of the selected user and emitting vocalizations corresponding to the first avatar using the trained voice model having been trained on the voice samples of the selected user.
R1. Examiner after reconsideration does not agree as it was found in paragraph 46 can be configured to analyze the audio including the voice communication, determine a reply (e.g., based on assigned characteristics or rules), and present the reply (e.g., by delivering pre-recorded or dynamic content generated through a text-to-speech engine), similar to the operation of a bot conversation. The mention of assigned characteristics or rules implies from paragraph 48 that recordings of the user are used to train an artificial intelligence model of how a user vocals output in order to clone the user.
“…As described above, the memory 202 can store software code representing instructions for the cloned avatar modeling and personalization module 213. In some implementations, the processor 201 can execute the instructions of the cloned avatar modeling and personalization module 213 to manage the generation and/or storing of characteristics of the cloned avatars that are related to the look and feel of each cloned avatar. Cloned avatars of a user can be modeled based on a primary avatar of a user and can optionally be personalized by the user (e.g., based on their respective passive or active roles and/or expected activities)…”
In paragraph 49 further teaches using the model to emulate voice output from a cloned avatar (emulate one or more user inputs that an online user would engage with when controlling the primary avatar of the user, such as by interacting… for example speaking into a microphone). Further based upon the modeled cloned avatar the system is able to emulate in passive or active mode to provide automated voice replies; these replies were modeled off of user audio inputs (voice samples per se).
“…A passive or active cloned avatar can provide automated voice replies to queries through, for example, a text-to-speech API, such that the cloned avatar communicates with other avatars similarly to when an online user speaks into a microphone and the recorded speech is associated with an avatar of the user (e.g., in real time) to communicate with another avatar…”
Paragraph 50 describes the storage solution to the audio/video capture systems 215 wherein frames captured are deconstructed into images and audio and analyzed separately which described in paragraph 51 this information is passed to the AI navigation and control systems module 216 to simulate behavior of the user through movement and voice outputs. The rest of paragraph 51 describes training the model that represents the users cloned avatar by analyzing the audio/video.
“…For example, after the control interface systems module 217 emulates a physical interaction between the user and a user input (e.g., a physical input controller), the AI navigation and control systems module 216 can analyze the outcome of a rendering engine associated with the virtual reality application (e.g., via analyzing data captured by the audio/video capture systems module 215) to determine an effect, if any, on and/or by the avatar resulting from the emulated physical interaction, and determining what, if anything, the cloned avatar should do next. For example, the analysis and determination by the AI navigation and control systems module 216 can be based on real-time inputs from, for example, the audio/video capture systems module 215. The AI navigation and control systems module 216 can send an instruction to the controller interface systems module 217 based on the determination of what the cloned avatar should do next so that the controller interface systems module 217 can emulate a physical interaction between the user and a user input to cause the cloned avatar to take a particular action (e.g., move, respond to another avatar, etc.)…”
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.
Inquires
Any inquiry concerning this communication should be directed to NICHOLAS AUGUSTINE at telephone number (571)270-1056.
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.
PNG
media_image1.png
213
559
media_image1.png
Greyscale
/NICHOLAS AUGUSTINE/Primary Examiner, Art Unit 2178 May 11, 2026
Prosecution Insights