Search

CN-121999805-A - Digital person control method, system and equipment for private interaction terminal

CN121999805ACN 121999805 ACN121999805 ACN 121999805ACN-121999805-A

Abstract

The embodiment of the application provides a digital person control method, a system and equipment for a private interaction terminal, which comprise the steps of obtaining a continuous signal value acquired by a sensor, generating a time sequence feature vector corresponding to the continuous signal value, determining an emotion tag corresponding to the time sequence feature vector, and an action tag and a voice expression parameter associated with the emotion tag, wherein the emotion tag represents the current emotion of a user, the action tag represents the action type which a digital person currently needs to execute, the voice expression parameter represents the voice intonation which the digital person currently needs to send, and driving the digital person to execute the corresponding facial expression, body action and voice intonation based on the emotion tag, the action tag and the voice expression parameter. The recognition capability of complex behavior states is provided, so that the change of digital human expression and actions is more natural.

Inventors

  • SONG HAO
  • WU DANNI

Assignees

  • 上海灵契数智科技有限公司

Dates

Publication Date
20260508
Application Date
20260123

Claims (10)

  1. 1. A digital person control method for a privacy interactive terminal, comprising: Acquiring continuous signal values acquired by a sensor, and generating time sequence feature vectors corresponding to the continuous signal values; Determining an emotion tag corresponding to the time sequence feature vector, and an action tag and a voice expression parameter associated with the emotion tag, wherein the emotion tag represents the current emotion of a user, the action tag represents the type of action which a digital person currently needs to execute, and the voice expression parameter represents the voice intonation which the digital person currently needs to send; And driving the digital person of the private interaction terminal to execute the corresponding facial expression, body action and voice intonation based on the emotion label, the action label and the voice expression parameter.
  2. 2. The digital person control method for a private interaction terminal according to claim 1, wherein the generating the timing feature vector corresponding to the continuous signal value includes: preprocessing the continuous signal values to obtain time sequence data, wherein the preprocessing comprises low-pass filtering processing, standardization processing and interval compression processing; and carrying out feature extraction on the change features in the time sequence data based on the set time window, and constructing corresponding time sequence feature vectors based on the feature extraction result, wherein the change features comprise one or more of pressure change features, acceleration change features, humidity change features, temperature change features and capacitance change features.
  3. 3. The digital person control method for a privacy interactive terminal according to claim 1, wherein the determining an emotion tag corresponding to the time series feature vector, and an action tag associated with the emotion tag, comprises: inputting the time sequence feature vector into a time sequence classification model trained by the private interaction terminal to obtain a corresponding emotion label, wherein the time sequence classification model is trained based on multidimensional change features; And acquiring interaction rules and current state change information of the digital person in the private interaction terminal, and determining an action tag according to the emotion tag, the interaction rules and the state change information.
  4. 4. The digital person control method for a private interaction terminal according to claim 3, wherein the determining an action tag according to the emotion tag, the interaction rule, and the state change information comprises: And determining a target action rule matched with the emotion label in a plurality of interaction rules, and adjusting action execution parameters in the target action rule based on the state change information to generate a corresponding action label, wherein the interaction rules are used for representing a reference execution action strategy corresponding to the corresponding emotion label.
  5. 5. The digital person control method for a privacy interactive terminal according to any one of claims 1 to 4, wherein the driving the digital person of the privacy interactive terminal to perform the corresponding facial expression, body motion, and voice intonation based on the emotion tag, the action tag, and the voice performance parameter comprises: Obtaining a corresponding face control parameter set according to the face mapping table which is inquired and set by the emotion tag, continuously interpolating and adjusting a parameter target interval corresponding to each expression parameter recorded in the face control parameter set, and controlling a digital person of the private interaction terminal to execute a corresponding face expression in real time based on an interpolation adjustment result, wherein the face control parameter set records parameter target intervals corresponding to a plurality of expression parameters; obtaining a corresponding playing mode and an action parameter range according to the action mapping table which is inquired and set by the action tag, and driving a digital person of the private interaction terminal to execute a corresponding body action according to the playing mode and the action parameter range; And determining voice performance parameters according to the emotion labels, adjusting any one or more of the voice speed, the pitch variation amplitude or the intonation fluctuation intensity of the digital person in the private interaction terminal according to the voice performance parameters, and playing the adjusted voice.
  6. 6. The digital person control method for a private interaction terminal according to any one of claims 1 to 4, wherein the generating a timing feature vector corresponding to the continuous signal value includes: generating time sequence feature vectors corresponding to the continuous signal values at intervals of a first preset time length; correspondingly, the driving the digital person of the private interaction terminal to execute the corresponding facial expression, body action and voice intonation based on the emotion label, the action label and the voice expression parameter comprises the following steps: And driving the digital person of the private interaction terminal to execute corresponding facial expression, body action and voice intonation based on the emotion label, the action label and the voice expression parameter every second preset time period, wherein the first preset time period is smaller than the second preset time period.
  7. 7. A digital personal control system for a privacy interactive terminal, comprising: The acquisition module is configured to acquire continuous signal values acquired by the sensor; The emotion judging module is configured to generate a time sequence feature vector corresponding to the continuous signal value, determine an emotion label corresponding to the time sequence feature vector, and an action label and a voice expression parameter associated with the emotion label, wherein the emotion label represents the current user emotion, the action label represents the action type which a digital person currently needs to execute, and the voice expression parameter represents the voice intonation which the digital person currently needs to send; and the engine module is configured to drive the digital person of the private interaction terminal to execute corresponding facial expression, body action and voice intonation based on the emotion label, the action label and the voice expression parameter.
  8. 8. A digital person control device for a privacy interactive terminal, characterized in that the digital person control device comprises one or more processors and a memory for storing one or more programs which, when executed by the one or more processors, cause the one or more processors to implement the digital person control method for a privacy interactive terminal as claimed in any one of claims 1 to 6.
  9. 9. A non-transitory computer-readable storage medium, characterized in that the non-transitory computer-readable storage medium stores therein a program or instructions that are executed by a processor to perform the digital human control method for a private interaction terminal according to any one of claims 1 to 7.
  10. 10. A computer program product comprising a computer program, characterized in that the computer program is stored in a non-volatile computer readable storage medium, from which at least one processor of a digital personal control device reads and executes the computer program, causing the digital personal control device to execute the digital personal control method for a private interaction terminal according to any of claims 1-6.

Description

Digital person control method, system and equipment for private interaction terminal Technical Field The embodiment of the application relates to the technical field of computers, in particular to a digital person control method, a digital person control system and digital person control equipment for a private interaction terminal. Background The development of the artificial intelligence technology of the virtual digital person is widely applied to various interaction scenes. Digital humans generally refer to digitized character figures created using digital technology that approximate human figures, with greater interactivity. For example, an application program for running a digital person is installed on the intelligent terminal, and the digital person is correspondingly displayed, and has interactive functions such as expression, action, voice and the like. In the related art, when a digital person performs interaction, a preset scenario script or a fixed event trigger based on a preset is adopted, so that the problem of single interaction mode exists. The emotion judgment is usually dependent on a single input, such as voice or buttons, and the like, does not have the recognition capability of complex behavior states, so that the interaction process of digital people, such as the change of expression and action, is unnatural, and a reasonable interaction control mechanism is lacked. Disclosure of Invention The embodiment of the application provides a digital person control method, a digital person control system and digital person control equipment for a private interaction terminal, which are used for determining emotion labels and action labels based on acquired continuous signal values of sensors and driving digital persons to execute corresponding facial expressions and body actions based on the emotion labels and the action labels, so that the problem of single interaction mode when digital person control is performed in the related technology is solved, and meanwhile, the digital person control system has the recognition capability of complex behavior states, so that the expression and action change is more natural. In a first aspect, an embodiment of the present application provides a digital person control method for a private interaction terminal, including: Acquiring continuous signal values acquired by a sensor, and generating time sequence feature vectors corresponding to the continuous signal values; Determining an emotion tag corresponding to the time sequence feature vector, and an action tag and a voice expression parameter associated with the emotion tag, wherein the emotion tag represents the current emotion of a user, the action tag represents the type of action which a digital person currently needs to execute, and the voice expression parameter represents the voice intonation which the digital person currently needs to send; And driving the digital person to execute corresponding facial expression, physical action and voice intonation based on the emotion label, the action label and the voice expression parameter. In a second aspect, an embodiment of the present application provides a digital person control system for a private interaction terminal, including: The acquisition module is configured to acquire continuous signal values acquired by the sensor; The emotion judging module is configured to generate a time sequence feature vector corresponding to the continuous signal value, determine an emotion label corresponding to the time sequence feature vector, and an action label and a voice expression parameter associated with the emotion label, wherein the emotion label represents the current user emotion, the action label represents the action type which a digital person currently needs to execute, and the voice expression parameter represents the voice intonation which the digital person currently needs to send; An engine module configured to drive the digital person to perform a corresponding facial expression, physical action, and speech intonation based on the emotion tags, the action tags, and the speech performance parameters. In a third aspect, an embodiment of the present application provides a digital person control apparatus for a private interaction terminal, where the apparatus includes one or more processors, and a memory configured to store one or more programs, which when executed by the one or more processors, cause the one or more processors to implement the digital person control method for a private interaction terminal according to the first aspect. In a fourth aspect, embodiments of the present application provide a non-volatile computer-readable storage medium having stored therein a program or instructions that is executed by a processor to perform the digital person control method for a privacy interaction terminal as set forth in any one of claims 1-7. In a fifth aspect, embodiments of the present application provide a computer program product comprisin