# US 12,118,652 B2 · Text-description based generation of avatars for artificial intelligence characters > Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt - Number: US 12,118,652 B2 - Document ID: US12118652B2 - Status: Granted - Kind: B2 - Application number: 18/401,376 - Filed: 2023-12-30 - Published: 2024-07-04 - Granted: 2024-10-15 - Earliest claimed: 2022-12-30 - Assignees: Theai, Inc. - Applicants: Theai, Inc. - Inventors: Ilya Gelfenbeyn; Mikhail Ermolenko; Kylan Gibbs; Aleksey Tikhonov - Web page: https://patentgrep.com/patent/US12118652B2 ## CPC - G06T 13/40: of characters, e.g. humans, animals or virtual beings - G06T 11/00: Two-dimensional [2D] image generation - G06T 17/00: Three-dimensional [3D] modelling for computer graphics - G06F 40/20: Natural language analysis (semantic analysis of natural language G06F40/30) - G06T: IMAGE DATA PROCESSING OR GENERATION, IN GENERAL - G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N) ## Abstract Systems and methods for text description-based generation of avatars for Artificial Intelligence (AI) characters are provided. An example method includes receiving a description of a face of an AI character, where the description is in a free-text format and where the face of the AI character is rendered by an AI character model in a virtual environment; acquiring at least one parameter of the AI character model, where the at least one parameter corresponds to at least one facial feature of the face of the AI character; analyzing the description to generate at least one value for the at least one parameter; and assigning the at least one value to the at least one parameter. ## Claims ### Claim 1 A method comprising: receiving, by at least one processor, a description of a face of an Artificial Intelligence (AI) character, the description being in a free-text format, the face of the AI character being rendered by an Al character model in a virtual environment, the AI character model including at least one parameter of the AI character model, the at least one parameter corresponding to at least one facial feature of the face of the AI character; generating, by the at least one processor, based on the description, at least one setting for the at least one parameter; tracking, by the at least one processor, interactions of the AI character with a user in the virtual environment; detecting, by the at least one processor, objects surrounding the AI character in a plurality of scenes in the virtual environment, wherein the AI character transitions from a first scene of the plurality of scenes to a second scene of the plurality of scenes; based on the interactions and the objects detected in the plurality of scenes, changing, by the at least one processor, runtime parameters of the AI character model, the runtime parameters being associated with an emotional state of the AI character; and during the interactions and based on the at least one setting and the runtime parameters, generating, by the at least one processor and using a pretrained machine learning model, a facial expression of the AI character. ### Claim 2 (depends on 1) The method of claim 1, wherein the at least one setting of the at least one parameter is determined by the pretrained machine learning model. ### Claim 3 (depends on 1) The method of claim 1, wherein the at least one facial feature includes one or more of the following: a type of a shape of the face of the AI character, a type of a shape of a body of the AI character, a color of the face, a length of hair of the AI character, a color of the hair, a ratio between a distance between eyes and a length of a nose, a color of eyes, a weight of the AI character, and a height of the AI character. ### Claim 4 (depends on 1) The method of claim 1, further comprising: determining that the description includes a name of a known fictional character or a known person; retrieving, based on the name, a further text description of the known fictional character or the known person from an external source; and adding the further text description to the description. ### Claim 5 (depends on 1) The method of claim 1, further comprising: determining that the description includes a first feature referenced to a first known fictional character or a first known person and a second feature referenced to a second known fictional character or a second known person; retrieving a first description of the first feature and a second description of the second feature from an external source; and adding the first description and the second description to the description. ### Claim 6 (depends on 1) The method of claim 1, wherein the AI character model includes a two-dimensional facial model configured to generate the face of the AI character. ### Claim 7 (depends on 1) The method of claim 1, wherein the AI character model includes a three-dimensional facial model configured to generate the face of the AI character. ### Claim 8 (depends on 1) The method of claim 1, wherein the AI character model is configured to: receive animation instructions; and animate the face of the AI character according to the animation instructions. ### Claim 9 (depends on 8) The method of claim 8, wherein the animation instructions are provided in the free-text format. ### Claim 10 (depends on 8) The method of claim 8, wherein the animation instructions include one or more of the following: a type of the facial expression of the AI character, a type of gesture of the AI character, and a type of movement of the AI character. ### Claim 11 A computing system comprising: a processor; and a memory storing instructions that, when executed by the processor, configure the computing system to: receive a description of a face of an Artificial Intelligence (AI) character, the description being in a free-text format, the face of the AI character being rendered by an AI character model in a virtual environment, the AI character model including at least one parameter of the AI character model, the at least one parameter corresponding to at least one facial feature of the face of the AI character; generate, based on the description, at least one setting for the at least one parameter; track interactions of the AI character with a user in the virtual environment; detect objects surrounding the AI character in a plurality of scenes in the virtual environment, wherein the AI character transitions from a first scene of the plurality of scenes to a second scene of the plurality of scenes; based on the interactions and the objects detected in the plurality of scenes, change runtime parameters of the AI character model, the runtime parameters being associated with an emotional state of the AI character; and during the interactions and based on the at least one setting and the runtime parameters, generate, using a pretrained machine learning model, a facial expression of the AI character. ### Claim 12 (depends on 11) The computing system of claim 11, wherein the at least one setting of the at least one parameter is determined by the pretrained machine learning model. ### Claim 13 (depends on 11) The computing system of claim 11, wherein the at least one facial feature includes one or more of the following: a type of a shape of the face of the AI character, a type of a shape of a body of the AI character, a color of the face, a length of hair of the AI character, a color of the hair, a ratio between a distance between eyes and a length of a nose, a color of eyes, a weight of the AI character, and a height of the AI character. ### Claim 14 (depends on 11) The computing system of claim 11, wherein the instructions further configure the computing system to: determine that the description includes a name of a known fictional character or a known person; retrieve, based on the name, a further text description of the known fictional character or the known person from an external source; and add the further text description to the description. ### Claim 15 (depends on 11) The computing system of claim 11, wherein the instructions further configure the computing system to: determine that the description includes a first feature referenced to a first known fictional character or a first known person and a second feature referenced to a second known fictional character or a second known person; retrieve a first description of the first feature and a second description of the second feature from an external source; and add the first description and the second description to the description. ### Claim 16 (depends on 11) The computing system of claim 11, wherein the AI character model includes a two-dimensional facial model configured to generate the face of the AI character. ### Claim 17 (depends on 11) The computing system of claim 11, wherein the AI character model includes a three-dimensional facial model configured to generate the face of the AI character. ### Claim 18 (depends on 11) The computing system of claim 11, wherein the AI character model is configured to: receive animation instructions; and animate the face of the AI character according to the animation instructions. ### Claim 19 (depends on 18) The computing system of claim 18, wherein the animation instructions are provided in the free-text format. ### Claim 20 A non-transitory computer-readable storage medium, the computer-readable storage medium including instructions that, when executed by a computing system, cause the computing system to: receive a description of a face of an Artificial Intelligence (AI) character, the description being in a free-text format, the face of the AI character being rendered by an AI character model in a virtual environment, the AI character model including at least one parameter of the AI character model, the at least one parameter corresponding to at least one facial feature of the face of the AI character; generate, based on the description, at least one setting for the at least one parameter; track interactions of the AI character with a user in the virtual environment; detect objects surrounding the AI character in a plurality of scenes in the virtual environment, wherein the AI character transitions from a first scene of the plurality of scenes to a second scene of the plurality of scenes; based on the interactions and the objects detected in the plurality of scenes, change runtime parameters of the AI character model, the runtime parameters being associated with an emotional state of the AI character; and during the interactions and based on the at least one setting and the runtime parameters, generate, using a pretrained machine learning model, a facial expression of the AI character.