Virtual classroom teaching simulation system and method based on natural interaction
Through the multimodal data collection and processing of the virtual classroom teaching simulation system, the spontaneous behavior and expression of virtual students are generated, which solves the problem that virtual classroom teaching cannot display real situations and language interactions, and improves teaching experience and teaching ability.
Patent Information
- Application Number
- CN202411289606.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-09-14
- Publication Date
- 2024-11-12
- Estimated Expiration
- Not applicable · inactive patent
AI Technical Summary
The existing virtual classroom teaching cannot intuitively and vividly display the real situation of classroom teaching, cannot achieve natural language interaction, there is a gap with the real classroom, and it has a poor sense of immersion.
Adopt a virtual classroom teaching simulation system based on natural interaction, including multimodal data acquisition module, preprocessing module, behavior control module, large language model module, environment control module and rendering module, through multimodal data acquisition, preprocessing, behavior The combined role of control, large language model and environmental control modules generates spontaneous behaviors and expressions of virtual students, realizes the movements and expression rendering of virtual students, and enhances the teaching experience.
It realizes the real situation of classroom teaching that can intuitively and vividly display, realizes natural language interaction, and improves the user's teaching ability and the richness of teaching scenarios.
Smart Images

Figure CN118941694A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of classroom teaching, and in particular to a virtual classroom teaching simulation system and method based on natural interaction. Background Art
[0002] The virtual classroom teaching simulation system is a comprehensive teaching platform based on computer technology, virtual reality technology and educational and teaching theory. It mainly provides a highly realistic and naturally interactive virtual classroom environment for normal school students or junior teachers, and cooperates with users to conduct classroom teaching drills to cultivate users' teaching ability and teacher-student interaction ability. At present, according to the different generation modes of virtual students, virtual classroom teaching platforms can be roughly divided into three categories: real-person control, computer program control and large language model control.
[0003] However, the existing virtual classroom teaching cannot intuitively and vividly display the real situation of classroom teaching, cannot achieve natural language interaction, has a gap with the real classroom, and has a poor sense of immersion. Summary of the invention
[0004] The purpose of the present invention is to provide a virtual classroom teaching simulation system and method based on natural interaction, aiming to solve the technical problems that the existing virtual classroom teaching cannot intuitively and vividly display the real situation of classroom teaching, cannot achieve natural language interaction, has a gap with the real classroom, and has a poor sense of immersion.
[0005] To achieve the above-mentioned purpose, the present invention adopts a virtual classroom teaching simulation system based on natural interaction, including a multimodal data acquisition module, a preprocessing module, a behavior control module, a large language model module, an environment control module and a rendering module, wherein the preprocessing module, the behavior control module and the large language module are all connected to the multimodal data acquisition module, the behavior control module and the large language module are also connected to the preprocessing module, the environment control module is connected to the preprocessing module, the behavior control module and the large language module, and the rendering module is connected to the environment control module;
[0006] The multimodal data acquisition module is used to collect audio and image information of the experimental environment and to collect the portrait settings of the virtual students;
[0007] The preprocessing module is used to preprocess the collected signal and convert it into a text signal to analyze the user's behavior intention;
[0008] The behavior control module is used to construct a large number of virtual students' action and expression libraries, and generate virtual students' spontaneous behaviors and expressions based on student portraits and classroom events;
[0009] The large language model module is used to generate answers to questions asked by users;
[0010] The environment control module is used to integrate the user's intention and the output results of the behavior control module and the large language model module;
[0011] The rendering module is used to render the actions and expressions of the virtual student and enable the virtual student to speak designated content.
[0012] Wherein, the multimodal data acquisition module includes a student portrait acquisition unit, an audio signal acquisition unit and a graphic signal acquisition unit, the student portrait acquisition unit is respectively connected to the behavior control module, the large language model module and the environment control module, and the audio signal acquisition unit and the graphic signal acquisition unit are both connected to the preprocessing module;
[0013] The student portrait collection unit is used to collect the portrait settings of the virtual students, and the student portrait collection unit is composed of a cognitive portrait and a personality portrait;
[0014] The audio signal acquisition unit is used to collect audio information of the experimental environment;
[0015] The graphic signal acquisition unit is used to acquire image information of the experimental environment.
[0016] Wherein, the preprocessing module includes a speech recognition model, an image analysis model and a semantic analysis model, the speech recognition model is connected to the audio signal acquisition unit, the image analysis model is connected to the graphic signal acquisition unit, and the semantic analysis model is connected to the speech recognition model, the image analysis model, the behavior control module, the large language model module and the environment control module respectively;
[0017] The speech recognition model is used to convert the original audio into text format data;
[0018] The image analysis model is used to extract and identify features of the input image, and also convert it into text format data;
[0019] The semantic analysis model is used to further perform semantic analysis on the text data output by the speech recognition model and the image analysis model to understand the user's intention.
[0020] The large language model module includes a memory unit and a large language model, the semantic analysis model and the student portrait acquisition unit are both connected to the memory unit, the memory unit is connected to the large language model, and the large language model is connected to the environment control module;
[0021] The large language model is responsible for the cognitive and thinking process, and generates intelligent answers to input questions in combination with the context;
[0022] The memory unit is used to simulate the memory ability of students.
[0023] Wherein, the environment control module includes a scene control unit and a student control unit, the scene control unit and the student control unit are both connected to the rendering module, the behavior control module is respectively connected to the scene control unit and the student control unit, the student portrait acquisition unit is respectively connected to the scene control unit and the student control unit, the large language model is respectively connected to the scene control unit and the student control unit, and the semantic analysis model is respectively connected to the scene control unit and the student control unit;
[0024] The scene control unit is used to receive a teacher's instruction to decide whether to switch the teaching scene;
[0025] The student control unit is used to simulate the behavior and state changes of the student.
[0026] Wherein, the rendering module includes an action rendering unit, an expression rendering unit and a text-to-speech model, and the action rendering unit, the expression rendering unit and the text-to-speech model are all connected to the scene control unit unit, and the action rendering unit, the expression rendering unit and the text-to-speech model are also connected to the student control unit;
[0027] The action rendering unit is used for rendering the virtual student actions;
[0028] The expression rendering unit is used for rendering the expression of the virtual student;
[0029] The text-to-speech model is used to convert the student's speech text into a speech file of a specified timbre and play it in real time.
[0030] The present invention also provides a virtual classroom teaching simulation method based on natural interaction, comprising the following steps:
[0031] The multimodal data acquisition module is used to collect audio and image information of the experimental environment and to collect the portrait settings of the virtual student;
[0032] The collected signal is preprocessed by the preprocessing module and converted into a text signal to analyze the user's behavior intention;
[0033] A large number of virtual students' action and expression libraries are constructed through the behavior control module, and the virtual students' spontaneous actions and expressions are generated according to the student portraits and classroom events;
[0034] Generate answers to user questions using the large language model module;
[0035] The environment control module integrates the user intention and the output results of the behavior control module and the large language model module;
[0036] The rendering module is used to render the actions and expressions of the virtual student, and enables the virtual student to speak designated content.
[0037] The present invention discloses a virtual classroom teaching simulation system and method based on natural interaction. The multimodal data acquisition module acquires audio and image information of an experimental environment and acquires portrait settings of virtual students. The preprocessing module preprocesses the acquired signals and converts them into text signals to analyze user behavior intentions. The behavior control module constructs a large number of virtual student action and expression libraries, generates virtual student spontaneous actions and expressions based on student portraits and classroom events, and provides users with more realistic classroom teaching experience and richer teaching scenes. The large language model module generates answers to user questions, and the environment control module integrates user intentions and output results of the behavior control module and the large language model module. The rendering module renders the actions and expressions of virtual students and enables the virtual students to speak designated content. Through the above-mentioned method, it is realized that the real situation of classroom teaching can be displayed intuitively and vividly, natural language interaction is realized, and the user's teaching ability can be quickly improved. BRIEF DESCRIPTION OF THE DRAWINGS
[0038] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.
[0039] Figure 1 It is a schematic diagram of the principle of the virtual classroom teaching simulation system based on natural interaction of the present invention.
[0040] Figure 2 It is a schematic diagram of the principle of the preprocessing module of the present invention.
[0041] Figure 3 It is a principle schematic diagram of the behavior control module of the present invention.
[0042] Figure 4 It is a schematic diagram of the principle of the large language model module of the present invention.
[0043] Figure 5 It is a principle schematic diagram of the environment control module of the present invention.
[0044] Figure 6 It is a rendering effect diagram of the rendering module of the present invention.
[0045] Figure 7 It is a schematic diagram of the principle of the virtual classroom teaching simulation method based on natural interaction of the present invention.
[0046] 100-multimodal data acquisition module, 200-preprocessing module, 300-behavior control module, 400-large language model module, 500-environment control module, 600-rendering module, 101-student portrait acquisition unit, 102-audio signal acquisition unit, 103-graphic signal acquisition unit, 201-speech recognition model, 202-image analysis model, 203-semantic analysis model, 401-memory unit, 402-large language model, 501-scene control unit, 502-student control unit, 601-action rendering unit, 602-expression rendering unit, 603-text-to-speech model. DETAILED DESCRIPTION
[0047] Embodiments of the present invention are described in detail below, examples of which are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are intended to be used to explain the present invention, and should not be construed as limiting the present invention.
[0048] See also Figure 1 to Figure 7 The present invention provides a virtual classroom teaching simulation system based on natural interaction, including a multimodal data acquisition module 100, a preprocessing module 200, a behavior control module 300, a large language model module 400, an environment control module 500 and a rendering module 600, wherein the preprocessing module 200, the behavior control module 300 and the large language module are all connected to the multimodal data acquisition module 100, the behavior control module 300 and the large language module are also connected to the preprocessing module 200, the environment control module 500 is connected to the preprocessing module 200, the behavior control module 300 and the large language module, and the rendering module 600 is connected to the environment control module 500;
[0049] The multimodal data acquisition module 100 is used to collect audio and image information of the experimental environment and to collect the portrait settings of the virtual student;
[0050] The preprocessing module 200 is used to preprocess the collected signal and convert it into a text signal to analyze the user's behavior intention;
[0051] The behavior control module 300 is used to construct a large number of virtual students' action and expression libraries, and to generate virtual students' spontaneous behaviors and expressions based on student portraits and classroom events;
[0052] The large language model module 400 is used to generate answers to questions asked by users;
[0053] The environment control module 500 is used to integrate the user's intention and determine the behavior and state changes of the environment and virtual students;
[0054] The rendering module 600 is used to render the actions and expressions of the virtual student, and use the text-to-speech model 603 to make the virtual student speak the specified content.
[0055] In this embodiment, the multimodal data acquisition module 100 acquires audio and image information of the experimental environment, and acquires the portrait setting of the virtual student. The preprocessing module 200 preprocesses the acquired signal and converts it into a text signal to analyze the user's behavioral intention. The behavior control module 300 constructs a large number of virtual students' action and expression libraries, generates virtual students' spontaneous behaviors and expressions based on student portraits and classroom events, and gives users a more realistic classroom teaching experience and richer teaching scenes. The large language model module 400 generates answers to user questions. The environment control module 500 integrates the user's intention and the output results of the behavior control module 300 and the large language model module 400. The rendering module 600 renders the virtual student's actions and expressions, and enables the virtual student to speak the specified content. Through the above method, it is possible to intuitively and vividly display the real situation of classroom teaching, realize natural language interaction, and quickly improve the user's teaching ability.
[0056] Further, the multimodal data acquisition module 100 includes a student portrait acquisition unit 101, an audio signal acquisition unit 102 and a graphic signal acquisition unit 103, the student portrait acquisition unit 101 is connected to the behavior control module 300, the large language model module 400 and the environment control module 500 respectively, and the audio signal acquisition unit 102 and the graphic signal acquisition unit 103 are both connected to the preprocessing module 200;
[0057] The student portrait collection unit 101 is used to collect the portrait settings of the virtual student, and the student portrait collection unit 101 is composed of a cognitive portrait and a personality portrait;
[0058] The audio signal acquisition unit 102 is used to collect audio information of the experimental environment;
[0059] The graphic signal acquisition unit 103 is used to acquire image information of the experimental environment.
[0060] In this embodiment, the student portrait acquisition unit 101 acquires the portrait settings of the virtual student. The student portrait acquisition unit 101 is composed of a cognitive portrait and a personality portrait. The audio signal acquisition unit 102 acquires audio information of the experimental environment. The graphic signal acquisition unit 103 acquires image information of the experimental environment. Among them, the student portrait is one of the most complex and critical inputs in the system. It defines the basic cognition and personality characteristics of the virtual student. The student portrait mainly consists of two parts: a cognitive portrait and a personality portrait.
[0061] Further, the preprocessing module 200 includes a speech recognition model 201, an image analysis model 202 and a semantic analysis model 203, the speech recognition model 201 is connected to the audio signal acquisition unit 102, the image analysis model 202 is connected to the graphic signal acquisition unit 103, and the semantic analysis model 203 is respectively connected to the speech recognition model 201, the image analysis model 202, the behavior control module 300, the large language model module 400 and the environment control module 500;
[0062] The speech recognition model 201 is used to convert the original audio into text format data;
[0063] The image analysis model 202 is used to extract and identify features of the input image and also convert it into text format data;
[0064] The semantic analysis model 203 is used to further perform semantic analysis on the text data output by the speech recognition model 201 and the image analysis model 202 to understand the user's intention.
[0065] In this embodiment, the speech recognition model 201 converts the original audio into text format data, the image analysis model 202 extracts and recognizes the input image features, and also converts it into text format data, and the semantic analysis model 203 further performs semantic analysis on the text data output by the speech recognition model and the image analysis model to understand the user's intention, and uses text classification technology to further abstract the user's semantic information, wherein the semantic information includes four categories: 1. Teacher instructions, which are used for discussion instructions, follow-up instructions, and classroom maintenance instructions; 2. Teacher lectures, which are used to teach teaching content; 3. Teacher status, teacher posture and expression information; 4. Teacher questions, questions that teachers expect virtual students to respond to. In the above method, complex multimodal input can be effectively converted into structured semantic understanding, providing accurate basic information for subsequent virtual student behavior and language generation. This multimodal fusion method not only improves the system's ability to understand the user's intention, but also can handle possible ambiguity or supplementary information between speech and action, thereby achieving more natural and intelligent human-computer interaction.
[0066] Further, the large language model module 400 includes a memory unit 401 and a large language model 402, the semantic analysis model 203 and the student portrait acquisition unit 101 are both connected to the memory unit 401, the memory unit 401 is connected to the large language model 402, and the large language model 402 is connected to the environment control module 500;
[0067] The large language model 402 is responsible for the cognitive and thinking process, and generates intelligent answers to input questions in combination with the context;
[0068] The memory unit 401 is used to simulate the memory ability of students.
[0069] In this embodiment, the large language model 402 is responsible for the cognitive and thinking process, and generates intelligent answers to input questions in combination with the context. The memory unit 401 simulates the student's memory ability, wherein the information stored in the memory unit 401 includes existing basic knowledge, customized information in the student portrait, and the content taught by the teacher in this class (the system will update the knowledge base in real time, and dynamically add the new knowledge points explained by the teacher in class to the memory unit 401). In the above method, it mainly provides "thinking" ability for virtual students. During classroom teaching, scenarios such as asking questions, discussing, mutual evaluation, and summarizing require virtual students to think about the specified content and give correct or wrong answers based on their own cognition.
[0070] Further, the environment control module 500 includes a scene control unit 501 and a student control unit 502, the scene control unit 501 and the student control unit 502 are both connected to the rendering module 600, the behavior control module 300 is respectively connected to the scene control unit 501 and the student control unit 502, the student portrait acquisition unit 101 is respectively connected to the scene control unit 501 and the student control unit 502, the large language model 402 is respectively connected to the scene control unit 501 and the student control unit 502, and the semantic analysis model 203 is respectively connected to the scene control unit 501 and the student control unit 502;
[0071] The scene control unit 501 is used to receive a teacher's instruction to decide whether to switch the teaching scene;
[0072] The student control unit 502 is used to control the behavior and state changes of the virtual student.
[0073] In this embodiment, the scene control unit 501 is used to receive the teacher's instructions to decide whether to switch the teaching scene, wherein the system defines multiple teaching scenes, including: teaching, discussion, mutual evaluation and sharing, etc. The module receives the teacher's instructions to decide whether to switch the scene, or enter the next stage of the scene; the student control unit 502 is used to simulate the behavior and state changes of the students, wherein the system defines multiple student action states, including: listening, answering, distraction, raising hands, standing up and discussing, etc. At the same time, multiple expressions are also defined, such as: confusion, smile, normal and surprise, etc. The student's actions and expressions are all converted and realized by the state transfer matrix, which determines whether the student state can be transferred and the conditions for the transfer.
[0074] Further, the rendering module 600 includes an action rendering unit 601, an expression rendering unit 602 and a text-to-speech model 603, and the action rendering unit 601, the expression rendering unit 602 and the text-to-speech model 603 are all connected to the scene control unit 501 unit, and the action rendering unit 601, the expression rendering unit 602 and the text-to-speech model 603 are also connected to the student control unit 502;
[0075] The action rendering unit 601 is used to render the actions of virtual students in different classes;
[0076] The expression rendering unit 602 is used to render the subtle facial expression changes of virtual students in different classes;
[0077] The text-to-speech model 603 is used to convert the student's speech text into a speech file with a specified timbre and play it in real time.
[0078] In this embodiment, the action rendering unit 601 is used to render the actions of the virtual students in different classes, and the expression rendering unit 602 is used to render the subtle facial expression changes of the virtual students in different classes.
[0079] The text-to-speech model 603 uses a neural network-based text-to-speech (TTS) model to convert the student's speech text into a voice file with a specified timbre and play it in real time.
[0080] The present invention also provides a virtual classroom teaching simulation method based on natural interaction, comprising the following steps:
[0081] S1: using the multimodal data acquisition module 100 to collect audio and image information of the experimental environment, and to collect the portrait settings of the virtual student;
[0082] S2: pre-processing the collected signal through the pre-processing module 200 and converting it into a text signal to analyze the user's behavior intention;
[0083] S3: constructing a large number of virtual students' action and expression libraries through the behavior control module 300, and generating virtual students' spontaneous actions and expressions based on student portraits and classroom events;
[0084] S4: Generate answers to user questions using the large language model module 400;
[0085] S5: The environment control module 500 integrates the user intention and the output results of the behavior control module 300 and the large language model module 400;
[0086] S6: Rendering the actions and expressions of the virtual student through the rendering module 600, and making the virtual student speak designated content.
[0087] In this embodiment, the multimodal data acquisition module 100 is first used to collect audio and image information of the experimental environment, and to collect the portrait setting of the virtual student. Then, the collected signal is preprocessed by the preprocessing module 200 and converted into a text signal to analyze the user's behavioral intention. A large number of virtual students' action and expression libraries are constructed by the behavior control module 300. The model is used to generate virtual students' spontaneous behaviors and expressions based on student portraits and classroom events. The large language model module 400 is used to generate answers to user questions. The environment control module 500 integrates the user's intention and the output results of the behavior control module 300 and the large language model module 400. The rendering module 600 renders the virtual student's actions and expressions, and enables the virtual student to speak the specified content.
[0088] What is disclosed above is only a preferred embodiment of the present invention, and it certainly cannot be used to limit the scope of rights of the present invention. Ordinary technicians in this field can understand that all or part of the processes of the above embodiment and equivalent changes made according to the claims of the present invention still fall within the scope of the invention.
Claims
1. A virtual classroom teaching simulation system based on natural interaction, characterized in that: It includes a multimodal data acquisition module, a preprocessing module, a behavior control module, a large language model module, an environment control module and a rendering module, wherein the preprocessing module, the behavior control module and the large language module are all connected to the multimodal data acquisition module, the behavior control module and the large language module are also connected to the preprocessing module, the environment control module is connected to the preprocessing module, the behavior control module and the large language module, and the rendering module is connected to the environment control module; The multimodal data acquisition module is used to collect audio and image information of the experimental environment and to collect the portrait settings of the virtual students; The preprocessing module is used to preprocess the collected signal and convert it into a text signal to analyze the user's behavior intention; The behavior control module is used to construct a large number of virtual students' action and expression libraries, and generate virtual students' spontaneous behaviors and expressions based on student portraits and classroom events; The large language model module is used to generate answers to questions asked by users; The environment control module is used to combine the user's intention and the output results of the behavior control module and the large language model module to determine the behavior and state changes of the environment and the virtual student; The rendering module is used to render the actions and expressions of the virtual student and enable the virtual student to speak designated content.
2. The virtual classroom teaching simulation system based on natural interaction as claimed in claim 1, characterized in that: The multimodal data acquisition module includes a student portrait acquisition unit, an audio signal acquisition unit and a graphic signal acquisition unit, the student portrait acquisition unit is connected to the behavior control module, the large language model module and the environment control module respectively, and the audio signal acquisition unit and the graphic signal acquisition unit are both connected to the preprocessing module; The student portrait collection unit is used to collect the portrait settings of the virtual students, and the student portrait collection unit is composed of a cognitive portrait and a personality portrait; The audio signal acquisition unit is used to collect audio information of the experimental environment; The graphic signal acquisition unit is used to acquire image information of the experimental environment.
3. The virtual classroom teaching simulation system based on natural interaction as claimed in claim 2, characterized in that: The preprocessing module includes a speech recognition model, an image analysis model and a semantic analysis model, the speech recognition model is connected to the audio signal acquisition unit, the image analysis model is connected to the graphic signal acquisition unit, and the semantic analysis model is connected to the speech recognition model, the image analysis model, the behavior control module, the large language model module and the environment control module respectively; The speech recognition model is used to convert the original audio into text format data; The image analysis model is used to extract and identify features of the input image, and also convert it into text format data; The semantic analysis model is used to further perform semantic analysis on the text data output by the speech recognition model and the image analysis model to understand the user's intention.
4. The virtual classroom teaching simulation system based on natural interaction as claimed in claim 3, characterized in that: The large language model module includes a memory unit and a large language model, the semantic analysis model and the student portrait acquisition unit are both connected to the memory unit, the memory unit is connected to the large language model, and the large language model is connected to the environment control module; The large language model is responsible for the cognitive and thinking process, and generates intelligent answers to input questions in combination with the context; The memory unit is used to simulate the memory ability of students.
5. The virtual classroom teaching simulation system based on natural interaction as claimed in claim 4, characterized in that: The environment control module includes a scene control unit and a student control unit, the scene control unit and the student control unit are both connected to the rendering module, the behavior control module is respectively connected to the scene control unit and the student control unit, the student portrait acquisition unit is respectively connected to the scene control unit and the student control unit, the large language model is respectively connected to the scene control unit and the student control unit, and the semantic analysis model is respectively connected to the scene control unit and the student control unit; The scene control unit is used to receive a teacher's instruction to decide whether to switch the teaching scene; The student control unit is used to simulate the behavior and state changes of the student.
6. The virtual classroom teaching simulation system based on natural interaction as claimed in claim 5, characterized in that: The rendering module includes an action rendering unit, an expression rendering unit and a text-to-speech model, wherein the action rendering unit, the expression rendering unit and the text-to-speech model are all connected to the scene control unit unit, and the action rendering unit, the expression rendering unit and the text-to-speech model are also connected to the student control unit; The action rendering unit is used for rendering the virtual student actions; The expression rendering unit is used for rendering the expression of the virtual student; The text-to-speech model is used to convert the student's speech text into a speech file of a specified timbre and play it in real time.
7. A virtual classroom teaching simulation method based on natural interaction, applied to the virtual classroom teaching simulation system based on natural interaction as claimed in claim 1, characterized in that: The steps include: The multimodal data acquisition module is used to collect audio and image information of the experimental environment and to collect the portrait settings of the virtual student; The collected signal is preprocessed by the preprocessing module and converted into a text signal to analyze the user's behavior intention; A large number of virtual students' action and expression libraries are constructed through the behavior control module, and the virtual students' spontaneous actions and expressions are generated according to the student portraits and classroom events; Generate answers to user questions using the large language model module; The environment control module integrates the user's intention and the output results of the behavior control module and the large language model module to determine the behavior and state changes of the environment and the virtual student; The rendering module is used to render the actions and expressions of the virtual student, and enables the virtual student to speak designated content.
Citation Information
Cited By
Virtual teaching training method and system based on large language model and medium
CN119624716A
Personalized teaching skill training system, device, medium and product based on large language model
CN119649657A
Personalized teaching skill training system, device, medium and product based on large language model
CN119649657B
