CN110866588A - Training learning method and system for realizing individuation of learnable ability model of intelligent virtual digital animal - Google Patents
Training learning method and system for realizing individuation of learnable ability model of intelligent virtual digital animal Download PDFInfo
- Publication number
- CN110866588A CN110866588A CN201911086750.4A CN201911086750A CN110866588A CN 110866588 A CN110866588 A CN 110866588A CN 201911086750 A CN201911086750 A CN 201911086750A CN 110866588 A CN110866588 A CN 110866588A
- Authority
- CN
- China
- Prior art keywords
- animal
- learnable
- model
- virtual digital
- intelligent virtual
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/004—Artificial life, i.e. computing arrangements simulating life
- G06N3/008—Artificial life, i.e. computing arrangements simulating life based on physical entities controlled by simulated intelligence so as to replicate intelligent life forms, e.g. based on robots replicating pets or humans in their appearance or behaviour
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F8/00—Arrangements for software engineering
- G06F8/60—Software deployment
- G06F8/65—Updates
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- General Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Software Systems (AREA)
- General Physics & Mathematics (AREA)
- Artificial Intelligence (AREA)
- Computing Systems (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Biophysics (AREA)
- Biomedical Technology (AREA)
- Life Sciences & Earth Sciences (AREA)
- Mathematical Physics (AREA)
- Health & Medical Sciences (AREA)
- Robotics (AREA)
- Computer Security & Cryptography (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses a training and learning method and a training and learning system for realizing the individuation of a learnable ability model of an intelligent virtual digital animal. The method comprises the following steps: generating a program example of the intelligent virtual digital animal at the cloud end; constructing a real object mapping of the production animal robot for the cloud example of the intelligent virtual digital animal, and sending the real object mapping to a user, wherein the animal robot downloads a corresponding learnable capability model from the cloud; the method comprises the following steps that a user carries out natural scene interaction with an animal robot object, the animal robot automatically collects, generates personalized training data, trains and updates a learnable ability model copy and uploads the learnable ability model copy to the cloud, and the cloud updates the learnable ability model of the corresponding virtual digital animal. The invention effectively solves the problems that the existing artificial intelligence model training mode has too high threshold for common people, cannot exert the value of mass labor force in a large range, cannot participate in the personalized training of the virtual digital intelligence capability model, has no real scene interaction with users and the like.
Description
Technical Field
The invention belongs to the technical field of machine learning training, and particularly relates to a training and learning method and system for realizing individuation of a learnable ability model of an intelligent virtual digital animal.
Background
The virtual digital animal is a virtual object which is defined by a basic attribute list and a behavior list through a computer program data structure and a code segment at a cloud server side, the virtual objects are usually concept mappings of one or more real animals in a real scene, and the concept mappings can even superpose attributes of other biological species and fictional species. With the development of artificial intelligence, people propose to add a learning capability model intelligent module which can be learned through training to a virtual digital animal, and the intelligent module is called as an intelligent virtual digital animal. The intelligent module of the learnable ability model can be represented by constructing a neural network model, and can realize training learning supporting various abilities by designing a plurality of intelligent modules for learning with different abilities for intelligent virtual digital animals.
In recent years, with the rapid advance of artificial intelligence technology theory and method and the great progress of intelligent manufacturing capability (the intelligence bodies such as robot dogs, robots, robot swarms, and the like are layered endlessly and updated), users have proposed the expectation of personalized training learning capability and interaction mode innovation (traditionally, interaction is completely performed in a software interface virtual scene, such as interaction with a virtual pet through a virtual graph and a picture interface drawn by software by using a mouse or a touch screen).
The technical method of the intelligent module for training the intelligent virtual digital animals at present is that a professional computer only trains a learnable model by using data collected uniformly in advance through professional means such as programming and the like to obtain a uniform capability model, and then the obtained uniform capability model is copied to user program object examples of each intelligent virtual digital animal indiscriminately. The shortcomings and disadvantages of the training technology, the method or the mode thereof include 'lack of personalized learning training of real interactive scene data of a user on a smart virtual digital animal' and 'lack of physical interaction between a person in a real scene and a virtual digital animal', which are specifically shown in the following steps: the user (common person) can not train and teach the intelligent virtual digital animal of the user individually and directly, and the user (common person) can only be in a software scene and can not directly interact with the intelligent virtual digital animal of the user in the physical world. It is difficult to satisfy the higher demands for the learning ability of the personalized training such as the smart digital pet dog partially or totally replacing the biological dog (the owner can selectively teach the biological dog to learn the personalized skill according to the needs of the owner), the smart virtual digital animal training which can be participated in by ordinary people without threshold like the domesticated animal, and the smart service robot dog, the smart service robot and the like which can trainably learn to grow according to the behavioral habits of a specific user.
Disclosure of Invention
The invention aims to: the method overcomes the defects of the prior training learning technology in the method mode, and provides a novel training learning method and a system for realizing the individuation of the learnable ability model of the intelligent virtual digital animal.
The invention manufactures a corresponding animal robot for each intelligent virtual digital animal, and the animal robot synchronously downloads an intelligent module of a learning capacity model of the intelligent virtual digital animal from a cloud and stores the intelligent module as a local copy. In a real scene that each user (common person) directly interacts with an animal robot belonging to the user, or in the process that the user consciously trains (educates, notices that the fixed parameters are different from the traditional configuration) the animal robot, personalized training data are automatically generated, a learnable ability model copy is trained, model parameters are updated and synchronized to the cloud, so that the problems of personalized learnable training of intelligent virtual digital animals by using real interaction scene data of the user lacking personalization, lack of physical interaction between people in the real scene and the virtual digital animals and the like are solved.
The technical scheme adopted by the invention is as follows:
in a first aspect, the present invention provides a training and learning method for realizing individuation of a learnable ability model of an intelligent virtual digital animal, comprising the steps of:
the animal robot real object downloads a copy of the learnable ability model from the cloud server side; the animal robot real object is an animal robot manufactured according to an intelligent virtual digital animal example at a cloud server end;
the animal robot real object performs scene interaction with a user to realize personalized training of a learnable ability model of the intelligent virtual digital animal;
the animal robot object uploads the trained learnable ability model copy to the cloud server, so that the cloud server can update the learnable ability model of the intelligent virtual digital animal according to the uploaded learnable ability model copy.
Further, the animal robot object downloads a copy of the learnable ability model from a cloud server, and the method includes:
the animal robot real object sends an identity verification request to a cloud server so that the cloud end can perform identity verification after receiving the request, and writes and locks a corresponding learnable capability model after the verification is passed;
the animal robot real object requests a cloud server end to download the learnable ability model;
verifying the data integrity of the downloaded learnable capability model by the animal robot real object;
and the animal robot real object loads the downloaded learnable ability model and sets the learnable ability model as an initial copy of the animal robot real object.
Further, the animal robot object performs scene interaction with a user to realize personalized training of a learnable ability model of the intelligent virtual digital animal, and the method comprises the following steps:
a user selects a training target for an animal robot real object;
carrying out personalized scene interaction between a user and an animal robot real object;
in the interaction process, the animal robot real object automatically acquires and generates personalized training data through a sensor, and the data are labeled according to a training target;
based on the collected data, the learnable ability model copy is updated by automatic training.
In a second aspect, the present invention provides a training learning method for realizing individuation of a learnable ability model of an intelligent virtual digital animal, which includes the steps of:
the cloud server side generates an intelligent virtual digital animal example;
the cloud server sends a learnable capability model copy to the animal robot real object according to a request of the animal robot real object; the animal robot real object is an animal robot manufactured according to an intelligent virtual digital animal example at a cloud server end;
the cloud server receives the learnable ability model copies uploaded by the animal robot entity and completed by training, and updates the learnable ability model of the intelligent virtual digital animal; the learnable ability model copy after training is a learnable ability model copy which is finished by a user through scene interaction with an animal robot object and personalized training.
Further, the cloud server side generates an intelligent virtual digital animal example by adopting the following steps:
registering a user;
the user selects the type of the intelligent virtual digital animal to be picked up;
an intelligent virtual digital animal program instance is generated for the user.
Further, the updating of the learnable capability model of the intelligent virtual digital animal includes:
verifying the data integrity of the learnable ability model copy;
releasing the write lock of the learnable ability model;
and updating the learnable capacity model of the intelligent virtual digital animal according to the uploaded learnable capacity model copy.
Furthermore, the cloud server side is provided with an intelligent virtual digital animal program object instance cloud hosting platform, the intelligent virtual digital animal program object instance cloud hosting platform runs program logic of an intelligent virtual digital animal at a cloud end, stores user resources, saves and manages virtual digital animal types, all attributes and behavior lists and attribute lists of the intelligent virtual digital animal, is responsible for interacting with animal robots in real objects, and provides access pages for users to directly access; the virtual digital animal type defines data structures of different animal types, including animal attributes, animal behaviors and a framework of a learnable ability model of the animal; the intelligent virtual digital animal is an instantiation program object of an intelligent virtual digital animal type and consists of a learnable ability model of the intelligent virtual digital animal, an attribute list and a behavior list, wherein the attribute list records attribute values of an intelligent virtual animal instance, and the behavior list defines basic action behaviors of the animal.
Further, the learnable ability model is composed of a learnable ability model structure definition, model parameters and a model parameter integrity check value; a set of parameter sets, namely model parameters, is obtained through training and learning, and an integrity check value is generated for each set of stable model parameters.
In a third aspect, the present invention provides an animal robot comprising a physical body and a computing system; the computing system downloads the learnable capability model copy of the intelligent virtual digital animal from the cloud server, achieves personalized training of the learnable capability model through scene interaction with a user, and uploads the trained learnable capability model copy to the cloud server, so that the cloud server can update the learnable capability model of the intelligent virtual digital animal according to the uploaded learnable capability model copy.
Furthermore, the animal robot is manufactured according to structural body information of an intelligent virtual digital animal example at a cloud server end, is a physical mapping of the intelligent virtual digital animal example, and realizes extension of a virtual digital world to a physical world; the physical body comprises physical presentation of attributes and behaviors of the intelligent virtual digital animal, the physical presentation comprises color, texture, a programmable control manipulator, a mechanical arm and a vision sensor, and the computing system comprises an instance management agent module, a learnable ability model copy module, an NPU module, a control logic integrated circuit and a storage module.
Further, the instance management agent module is communicated with an instance management module of the cloud server through a network and is responsible for submitting an identity authentication request to the cloud hosting platform, downloading a learnable capability model, checking the downloaded learnable capability model, installing a local model copy of the learnable capability model as an animal robot real object, scheduling, training and updating a program of the learnable capability model copy, and uploading the updated learnable capability model copy to the cloud; the NPU module is an intelligent computing special processor and is used for accelerating the neural network reasoning and training performance, the control logic integrated circuit is a processor for realizing programmable control of the animal robot, and the storage module provides a rapid storage function for a learnable ability model copy, training data and program operation.
In a fourth aspect, the present invention provides a cloud server provided with an intelligent virtual digital animal program object instance cloud hosting platform;
the intelligent virtual digital animal program object instance cloud hosting platform generates an intelligent virtual digital animal instance, and a learnable capability model copy is sent to the animal robot entity according to a request of the animal robot entity; the animal robot real object is an animal robot manufactured according to an intelligent virtual digital animal example at a cloud server end;
the intelligent virtual digital animal program object instance cloud hosting platform receives a learnable ability model copy uploaded by an animal robot real object and completed by training, and updates the learnable ability model of the intelligent virtual digital animal; the learnable ability model copy after training is a learnable ability model copy which is finished by a user through scene interaction with an animal robot object and personalized training.
Further, the cloud server comprises an instance management module, a virtual digital animal type module and an intelligent virtual animal instance module;
the instance management module manages the full life cycle of the intelligent virtual digital animal instance and the learnable capability model management of the instance, and is responsible for carrying out interactive communication with the instance management agent module running in the animal robot real object through a network;
the virtual digital animal type module comprises different animal type types, defines data structure bodies of different animal types and comprises animal attributes, animal behaviors and a framework of a model of the learnable ability of the animal;
the intelligent virtual animal instance module is a concrete instantiation of an intelligent virtual digital animal type and comprises a learnable ability model, an attribute list and a behavior list.
In a fifth aspect, the invention provides a training learning system for realizing personalization of a learnable ability model of an intelligent virtual digital animal, which comprises the animal robot and a cloud server.
Compared with the prior art, the invention has the advantages that:
(1) at present, training of artificial intelligence capability models (such as RNN, CNN network, deep reinforcement model and the like) of intelligent virtual digital animals requires professional computer engineers, and the threshold is high; the invention provides a method and a mode for training a capability model for an intelligent virtual digital animal in the natural scene interaction of a common person and an offline animal robot real object for the on-line intelligent virtual digital animal production line animal robot real object mapping, so that the threshold of the artificial intelligent capability model for training the virtual digital animal is obviously reduced, and the new technical mode provides a simple and convenient way for mass labor force to participate in the capability model training of a virtual digital intelligent agent.
(2) Currently, the learnable ability model training of intelligent virtual digital animals adopts the ways of uniformly collecting data, uniformly training and uniformly deploying on line or before the online, and has the problems of lack of personalized intelligence, flexibility and real scene participation sense; the training learning method and the training learning system are very flexible, and the problems of individuation of a learnable ability model of intelligent virtual digital animals for a large number of users, physical interaction of real scenes of people and the virtual digital animals and the like are solved. The training and interaction mode combining the online intelligent virtual digital instance and the offline real robot mapping provides a technical method and a mode for realizing the application of digital pets and the like which continuously learn to grow and have personalized capabilities (such as learning to call, identify colors, pick up thrown balls and the like) under the training of an owner.
Drawings
FIG. 1 is a schematic view of the overall structure of the present invention;
FIG. 2 is a process of the present invention for generating an animal robot map for an intelligent virtual digital animal construction;
fig. 3 is a training learning process of the present invention implementing the intelligent virtual digital animal's learnability model personalization.
Detailed Description
The invention is further illustrated with reference to the following figures and examples, without in any way limiting the scope of the invention.
The training and learning system for realizing the individuation of the learnable ability model of the intelligent virtual digital animal is mainly composed of an intelligent virtual digital animal program object instance cloud hosting platform 1, an animal robot real object 2 and a network 3, as shown in fig. 1. The cloud hosting platform 1 for the intelligent virtual digital animal program object instance is mainly composed of an instance management module 101, a virtual digital animal type module 102, an intelligent virtual digital animal instance module 103 (a plurality of different instances form an instance set), and the like. Intelligent virtual digital animal instance module 103 contains a learnable capability model 1031, an attribute list 1032, and a behavior list 1033. The animal robot real object 2 is composed of an instance management agent module 201, a learnable ability model copy module 202, an NPU (neural network processor) module 203, a storage module 205, a control logic integrated circuit 204, a physical body and the like. The blocks shown in the figure by dashed boxes are not a core component of the present invention. The person 4 is a user, is an owner of an intelligent virtual digital animal example and a corresponding animal robot real object, directly interacts with the corresponding animal robot real object in a physical scene, the animal robot real object acquires and generates training data in interaction through a visual sensor, a collision detection sensor, a touch sensor, a pressure sensor, an optical fiber sensor and the like, and the learnable ability model copy is trained and updated. The intelligent virtual digital animal program object instance cloud hosting platform 1 runs at a cloud end, the animal robot real object 2 is at a user end (a common person), and the network 3 is connected with the cloud end and the animal robot real object and provides a data transmission link.
The cloud hosting platform 1 for the intelligent virtual digital animal program object instance runs an intelligent virtual digital animal program logic at the cloud end, stores user resources, stores and manages virtual digital animal types and all attribute and behavior attribute lists of intelligent virtual digital animals, is responsible for physical interaction with an animal robot, and also provides an access page for users (people) to directly access and the like. The instance management module 101 manages the full life cycle of the intelligent virtual digital animal instance, the learnable ability model management of the instance (locking the learnable ability model, unlocking the learnable ability model, updating the learnable ability model, etc.), and is responsible for interactive communication with the instance management agent module 201 running in the animal robot real object 2 through the network 3, and the communication contents mainly include verification of user and instance identities, provision of a learnable ability model download, reception of a learnable ability model upload, verification of a learnable ability model, etc.
The virtual digital animal type module 102 is a framework of different animal type types managed by the platform, and defines data structures of different animal types, such as animal attributes, animal behaviors, and learnable ability models of animals.
The intelligent virtual animal instance module 103 is a concrete instantiation of intelligent virtual digital animal types, each animal type generally has a plurality of instances, each instance has an independently stored learnable capability model, and all the instances form an instance set. The smart virtual animal instance is composed of a learnable ability model 1031, an attribute list 1032, a behavior list 1033, and the like. Wherein the attribute list defines attributes (such as unique identification ID, color, age, gender and the like) of the intelligent virtual animal instance, and the behavior list defines basic action behaviors (such as running, walking, turning, jumping, vigorous calling and the like) of the animal.
The learnable model 1031 is composed of a learnable model structure definition, model parameters, and a model parameter integrity check value, where the learnable model structure may be an Artificial Neural Network (ANN), a deep reinforcement learning (Reinforcement learning), a spiking neural network (SNn), and the like, and different types of the learnable model structure do not affect the rights of the present invention. A group of parameter sets, namely model parameters, is obtained through training and learning, and for each group of stable model parameters, a hash value is generated to serve as an integrity check value (an integrity check value generation algorithm is optional, and other algorithms except the hash algorithm are selected and do not influence the right of the invention).
The animal robot real object 2 is an animal robot manufactured according to the structural body information of the intelligent virtual digital animal example module 103, is a physical mapping of an intelligent virtual digital animal example, and realizes extension of a virtual digital world to a physical world. The method and the mode for training the intelligent virtual digital animal are firstly proposed by the method, and are a key step for realizing that common public users (but not professional computer users) can purposefully generate data and train models in a real interactive scene in combination with personal desires to achieve the personalized ability training of the intelligent virtual digital animal. The animal robot real object consists of a physical body and a computing system. The physical body comprises physical presentation of attributes and behaviors of the intelligent virtual digital animal, such as color, texture, a programmable control manipulator, a mechanical arm, a visual sensor and the like; the design, implementation and manufacturing method of the physical bodies are not important components of the invention, and the different types of the manufacturing method, the technology, the process and the like do not influence the right of the invention. The computing system is a key point and mainly comprises an instance management agent module 201, a learnable ability model copy module 202, an NPU module 203, a control logic integrated circuit 204 and a storage module 205.
The instance management agent module 201 is in communication with the instance management module 101 through the network 3, and is responsible for submitting an identity authentication request to the cloud hosting platform, downloading the learnable capability model 1031, verifying the downloaded learnable capability model 1031, installing the learnable capability model 1031 as a local copy, and uploading the learnable capability model copy 202 updated by the user terminal training to the cloud. The NPU module 203 is an intelligent computing special processor and is used for accelerating neural network reasoning and training performance, the control logic integrated circuit 204 is a processor for realizing programmable control of the animal robot, the storage module 205 provides rapid storage for learnable ability model copies, training data, program operation and the like, and different types of the modules 203, 204 and 205 do not influence the right of the invention.
The present invention will be further described with reference to fig. 2 and 3 by two general users who respectively train their own intelligent virtual digital dogs to learn different skills. Fig. 2 illustrates a process of generating an animal robot real object map for intelligent virtual digital animal construction, and fig. 3 illustrates a training learning process for realizing the individualization of a learnable ability model of an intelligent virtual digital animal.
(1) An animal robot physical object map is generated for the intelligent virtual digital animal construction. As shown in fig. 2, the method mainly comprises the following steps:
(1.1) registering an intelligent virtual digital animal program object instance cloud hosting platform 1 by a user A, and picking up an intelligent virtual digital animal instance; in this embodiment, the user a obtains the intelligent virtual digital dog, and the attribute list thereof includes a unique identity number of 10001, a color of gray, an age of 3 months, and the like; the model framework of the model of 10001 intelligent virtual digital dog, such as image recognition, voice understanding, motion trail tracking and the like, is integrated, and therefore the model framework has the capabilities of obtaining recognized images, understanding voice, automatically adjusting motion trail and the like through training and learning. Wherein the image recognition model is implemented based on a CNN network (convolutional neural network);
(1.2) producing a physical map of an animal robot for manufacturing a virtual digital dog No. 10001;
(1.3) distributing the animal robot object No. 10001 to a user A;
(1.4) starting the No. 10001 animal robot real object by the user A, and networking the robot;
(1.5) sending an identity verification request to a cloud hosting platform by the animal robot No. 10001;
(1.6) the cloud instance management module receives the request, performs identity verification, identifies the request as a 10001 animal robot physical object (if identity verification fails, identity verification failure information is fed back to the animal robot physical object, and the animal robot physical object can resend the identity verification request), and then performs write locking on the learnable capacity model of the 10001 intelligent virtual digital animal (so as to avoid model inconsistency or write conflict caused by parallel write);
(1.7) No. 10001 animal robot requests to download a model with learnable ability of No. 10001 intelligent virtual digital animal hosted by a cloud, the downloading is completed, and model integrity check is passed (if the integrity check is not passed, the model downloading is not complete, the model downloading needs to be downloaded again);
(1.8) loading the downloaded model, and initializing to be a copy of the learnable capability model of the animal robot physical object No. 10001.
(2) And carrying out personalized training on the learnable capability model of the intelligent virtual digital animal. As shown in fig. 3, the method mainly comprises the following steps:
(2.1) the user A selects a training target through a 10001 animal robot real object (which can be realized by voice awakening, menu selection and other technologies, and the mode is not limited): learning to identify apples; then the robot real object enters a training interaction mode;
(2.2) the user A displays different apples and different sides of the apples in front of the animal robot No. 10001, the animal robot collects photos of the apples through a visual sensor, and the acquisition of new photos is triggered by the movement and angle change of the apples each time;
(2.3) in combination with the training target set in the step 2.1), all the collected photos are labeled as 'apples', the animal robot No. 10001 uses NPU to accelerate, and trains and updates the image classification subsystem model of the learnable capability model copy until convergence. The image classification subsystem is realized on the basis of technologies such as a CNN convolutional neural network and a multi-classifier (but other technology choices do not influence the right of the invention);
(2.4) the 10001 animal robot uploads the learnable ability model copy after training and updating to the cloud;
(2.5) the cloud instance management module receives the uploading and verifies the integrity of the model data (if the integrity is not verified, the model is not uploaded completely, and the model needs to be uploaded again);
(2.6) the cloud instance management module releases the write-locking of the learnable capacity model of the intelligent virtual digital animal instance No. 10001;
and (2.7) updating the learnable capability model of the virtual digital animal No. 10001 by the cloud according to the uploaded capability model copy. Therefore, the intelligent virtual digital animal of the user A has the capability of identifying the apple in a personalized mode. In the training process, the user A generates personalized training data through the scene interaction side model, and is suitable for the common public.
The unique identity number of the intelligent virtual digital dog received by the user B is 10002, the user B can train the 10002 intelligent virtual digital dog to obtain the capability of identifying individuation of football, basketball, orange and the like through a process similar to that of the user A, and common users can train the intelligent virtual digital animal to obtain the track tracking capability of a specific form and other learning capabilities by adopting a similar training method mode. Under the low-technology threshold training method and the low-technology threshold training mode, the intelligent virtual digital dogs of the users A and B show differentiated and personalized trainable learning ability 'growth'.
The invention provides a novel model training technical mode for realizing the training and learning method and the system for realizing the individuation of the learnable ability model of the intelligent virtual digital animal, and provides a simple, convenient and efficient way for the common public labor force to participate in the ability model training of the virtual digital intelligent agent.
In the specific implementation of the scheme of the invention, after a user helps an animal robot object to generate or collect personalized training data through scene interaction in the training process, if the system does not immediately perform model training at the animal robot object end, the data is uploaded to the cloud end, and the learning capability model of the corresponding virtual digital animal of the user is updated by being postponed to the cloud end training, so that the personalization of the learning capability model is realized, which is regarded as a deformation mode of the invention.
The above embodiments are only intended to illustrate the technical solution of the present invention and not to limit the same, and a person skilled in the art can modify the technical solution of the present invention or substitute the same without departing from the principle and scope of the present invention, and the scope of the present invention should be determined by the claims.
Claims (14)
1. A training and learning method for realizing the individuation of a learnable ability model of an intelligent virtual digital animal comprises the following steps:
the animal robot real object downloads a copy of the learnable ability model from the cloud server side; the animal robot real object is an animal robot manufactured according to an intelligent virtual digital animal example at a cloud server end;
the animal robot real object performs scene interaction with a user to realize personalized training of a learnable ability model of the intelligent virtual digital animal;
the animal robot object uploads the trained learnable ability model copy to the cloud server, so that the cloud server can update the learnable ability model of the intelligent virtual digital animal according to the uploaded learnable ability model copy.
2. The method of claim 1, wherein the downloading of the copy of the learnable capability model from the cloud server by the animal robot entity comprises:
the animal robot real object sends an identity verification request to a cloud server so that the cloud end can perform identity verification after receiving the request, and writes and locks a corresponding learnable capability model after the verification is passed;
the animal robot real object requests a cloud server end to download the learnable ability model;
verifying the data integrity of the downloaded learnable capability model by the animal robot real object;
and the animal robot real object loads the downloaded learnable ability model and sets the learnable ability model as an initial copy of the animal robot real object.
3. The method of claim 1, wherein the animal robot object performs scene interaction with a user to realize personalized training of a learnable capability model of the intelligent virtual digital animal, and the method comprises the following steps:
a user selects a training target for an animal robot real object;
carrying out personalized scene interaction between a user and an animal robot real object;
in the interaction process, the animal robot real object automatically acquires and generates personalized training data through a sensor, and the data are labeled according to a training target;
based on the collected data, the learnable ability model copy is updated by automatic training.
4. A training and learning method for realizing the individuation of a learnable ability model of an intelligent virtual digital animal comprises the following steps:
the cloud server side generates an intelligent virtual digital animal example;
the cloud server sends a learnable capability model copy to the animal robot real object according to a request of the animal robot real object; the animal robot real object is an animal robot manufactured according to an intelligent virtual digital animal example at a cloud server end;
the cloud server receives the learnable ability model copies uploaded by the animal robot entity and completed by training, and updates the learnable ability model of the intelligent virtual digital animal; the learnable ability model copy after training is a learnable ability model copy which is finished by a user through scene interaction with an animal robot object and personalized training.
5. The method according to claim 4, wherein the cloud server generates the intelligent virtual digital animal instance by adopting the following steps:
registering a user;
the user selects the type of the intelligent virtual digital animal to be picked up;
an intelligent virtual digital animal program instance is generated for the user.
6. The method of claim 4, wherein updating the learnable capability model of the intelligent virtual digital animal comprises:
verifying the data integrity of the learnable ability model copy;
releasing the write lock of the learnable ability model;
and updating the learnable capacity model of the intelligent virtual digital animal according to the uploaded learnable capacity model copy.
7. The method according to claim 4, wherein the cloud server side is provided with an intelligent virtual digital animal program object instance cloud hosting platform, the intelligent virtual digital animal program object instance cloud hosting platform runs program logic of an intelligent virtual digital animal in a cloud end, stores user resources, saves and manages virtual digital animal types and all attributes and behavior lists and attribute lists of the intelligent virtual digital animal, is responsible for interacting with animal robots and provides access pages for users to directly access; the virtual digital animal type defines data structures of different animal types, including animal attributes, animal behaviors and a framework of a learnable ability model of the animal; the intelligent virtual digital animal is an instantiation program object of an intelligent virtual digital animal type and consists of a learnable ability model of the intelligent virtual digital animal, an attribute list and a behavior list, wherein the attribute list records attribute values of an intelligent virtual animal instance, and the behavior list defines basic action behaviors of the animal.
8. The method of claim 4, wherein the learnable capability model consists of a learnable capability model structure definition, model parameters, and model parameter integrity check values; a set of parameter sets, namely model parameters, is obtained through training and learning, and an integrity check value is generated for each set of stable model parameters.
9. An animal robot, comprising a physical body and a computing system; the computing system downloads the learnable capability model copy of the intelligent virtual digital animal from the cloud server, achieves personalized training of the learnable capability model through scene interaction with a user, and uploads the trained learnable capability model copy to the cloud server, so that the cloud server can update the learnable capability model of the intelligent virtual digital animal according to the uploaded learnable capability model copy.
10. The animal robot of claim 9, wherein the animal robot is manufactured according to structural information of an intelligent virtual digital animal instance at a cloud server, is a physical mapping of the intelligent virtual digital animal instance, and realizes extension of a virtual digital world to a physical world; the physical body comprises physical presentation of attributes and behaviors of the intelligent virtual digital animal, the physical presentation comprises color, texture, a programmable control manipulator, a mechanical arm and a vision sensor, and the computing system comprises an instance management agent module, a learnable ability model copy module, an NPU module, a control logic integrated circuit and a storage module.
11. The animal robot of claim 10, wherein the instance management agent module communicates with the instance management module of the cloud server via a network, and is responsible for submitting an identity authentication request to the cloud hosting platform, downloading a learnable capability model, verifying the downloaded learnable capability model, installing a local model copy of the learnable capability model as an animal robot object, scheduling, training and updating a program of the learnable capability model copy, and uploading the updated learnable capability model copy to the cloud; the NPU module is an intelligent computing special processor and is used for accelerating the neural network reasoning and training performance, the control logic integrated circuit is a processor for realizing programmable control of the animal robot, and the storage module provides a rapid storage function for a learnable ability model copy, training data and program operation.
12. A cloud server is characterized in that an intelligent virtual digital animal program object instance cloud hosting platform is arranged;
the intelligent virtual digital animal program object instance cloud hosting platform generates an intelligent virtual digital animal instance, and a learnable capability model copy is sent to the animal robot entity according to a request of the animal robot entity; the animal robot real object is an animal robot manufactured according to an intelligent virtual digital animal example at a cloud server end;
the intelligent virtual digital animal program object instance cloud hosting platform receives a learnable ability model copy uploaded by an animal robot real object and completed by training, and updates the learnable ability model of the intelligent virtual digital animal; the learnable ability model copy after training is a learnable ability model copy which is finished by a user through scene interaction with an animal robot object and personalized training.
13. The cloud server of claim 12, comprising an instance management module, a virtual digital animal type module, and a smart virtual animal instance module;
the instance management module manages the full life cycle of the intelligent virtual digital animal instance and the learnable capability model management of the instance, and is responsible for carrying out interactive communication with the instance management agent module running in the animal robot real object through a network;
the virtual digital animal type module comprises different animal type types, defines data structure bodies of different animal types and comprises animal attributes, animal behaviors and a framework of a model of the learnable ability of the animal;
the intelligent virtual animal instance module is a concrete instantiation of an intelligent virtual digital animal type and comprises a learnable ability model, an attribute list and a behavior list.
14. A training learning system for realizing personalization of a learnable ability model of an intelligent virtual digital animal, comprising the animal robot of any one of claims 9 to 11 and the cloud server of claim 12 or 13.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201911086750.4A CN110866588B (en) | 2019-11-08 | 2019-11-08 | Training learning method and system for realizing individuation of learning ability model of intelligent virtual digital animal |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201911086750.4A CN110866588B (en) | 2019-11-08 | 2019-11-08 | Training learning method and system for realizing individuation of learning ability model of intelligent virtual digital animal |
Publications (2)
Publication Number | Publication Date |
---|---|
CN110866588A true CN110866588A (en) | 2020-03-06 |
CN110866588B CN110866588B (en) | 2023-05-02 |
Family
ID=69653228
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201911086750.4A Active CN110866588B (en) | 2019-11-08 | 2019-11-08 | Training learning method and system for realizing individuation of learning ability model of intelligent virtual digital animal |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN110866588B (en) |
Cited By (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111461694A (en) * | 2020-03-17 | 2020-07-28 | 上海大学 | Dish identification and pricing system and method based on multi-level learning model |
CN112214748A (en) * | 2020-10-30 | 2021-01-12 | 支付宝(杭州)信息技术有限公司 | Identity recognition system, method and device |
CN112494956A (en) * | 2020-12-15 | 2021-03-16 | 四川长虹电器股份有限公司 | Simulation method and simulation system for converting articles into pets |
CN112668687A (en) * | 2020-12-01 | 2021-04-16 | 达闼机器人有限公司 | Cloud robot system, cloud server, robot control module and robot |
CN113435126A (en) * | 2021-07-07 | 2021-09-24 | 魏天骐 | Knowledge sharing processing method, intelligent robot equipment, knowledge sharing system and task learning system |
WO2022021421A1 (en) * | 2020-07-31 | 2022-02-03 | Oppo广东移动通信有限公司 | Model management method, system and apparatus, communication device, and storage medium |
Citations (12)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101174332A (en) * | 2007-10-29 | 2008-05-07 | 张建中 | Method, device and system for interactively combining real-time scene in real world with virtual reality scene |
US20090090305A1 (en) * | 2007-10-03 | 2009-04-09 | National University Of Singapore | System for humans and pets to interact remotely |
CA2941250A1 (en) * | 2014-03-19 | 2015-09-24 | Neurala, Inc. | Methods and apparatus for autonomous robotic control |
CN107139179A (en) * | 2017-05-26 | 2017-09-08 | 西安电子科技大学 | A kind of intellect service robot and method of work |
DE202017106132U1 (en) * | 2016-10-10 | 2017-11-13 | Google Llc | Neural networks for selecting actions to be performed by a robot agent |
WO2018089700A1 (en) * | 2016-11-10 | 2018-05-17 | Warner Bros. Entertainment Inc. | Social robot with environmental control feature |
CN108897848A (en) * | 2018-06-28 | 2018-11-27 | 北京百度网讯科技有限公司 | Robot interactive approach, device and equipment |
WO2019057019A1 (en) * | 2017-09-20 | 2019-03-28 | 阿里巴巴集团控股有限公司 | Robot interaction method and device |
US20190102684A1 (en) * | 2017-09-29 | 2019-04-04 | Sony Interactive Entertainment Inc. | Mobile and autonomous personal companion based on an artificial intelligence (ai) model for a user |
CN109656529A (en) * | 2018-10-31 | 2019-04-19 | 北京大学 | A kind of on-line customization method and system for client deep learning |
CN110298212A (en) * | 2018-03-21 | 2019-10-01 | 腾讯科技(深圳)有限公司 | Model training method, Emotion identification method, expression display methods and relevant device |
CN110378463A (en) * | 2019-07-15 | 2019-10-25 | 北京智能工场科技有限公司 | A kind of artificial intelligence model standardized training platform and automated system |
-
2019
- 2019-11-08 CN CN201911086750.4A patent/CN110866588B/en active Active
Patent Citations (13)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20090090305A1 (en) * | 2007-10-03 | 2009-04-09 | National University Of Singapore | System for humans and pets to interact remotely |
CN101174332A (en) * | 2007-10-29 | 2008-05-07 | 张建中 | Method, device and system for interactively combining real-time scene in real world with virtual reality scene |
CA2941250A1 (en) * | 2014-03-19 | 2015-09-24 | Neurala, Inc. | Methods and apparatus for autonomous robotic control |
US20170024877A1 (en) * | 2014-03-19 | 2017-01-26 | Neurala, Inc. | Methods and Apparatus for Autonomous Robotic Control |
DE202017106132U1 (en) * | 2016-10-10 | 2017-11-13 | Google Llc | Neural networks for selecting actions to be performed by a robot agent |
WO2018089700A1 (en) * | 2016-11-10 | 2018-05-17 | Warner Bros. Entertainment Inc. | Social robot with environmental control feature |
CN107139179A (en) * | 2017-05-26 | 2017-09-08 | 西安电子科技大学 | A kind of intellect service robot and method of work |
WO2019057019A1 (en) * | 2017-09-20 | 2019-03-28 | 阿里巴巴集团控股有限公司 | Robot interaction method and device |
US20190102684A1 (en) * | 2017-09-29 | 2019-04-04 | Sony Interactive Entertainment Inc. | Mobile and autonomous personal companion based on an artificial intelligence (ai) model for a user |
CN110298212A (en) * | 2018-03-21 | 2019-10-01 | 腾讯科技(深圳)有限公司 | Model training method, Emotion identification method, expression display methods and relevant device |
CN108897848A (en) * | 2018-06-28 | 2018-11-27 | 北京百度网讯科技有限公司 | Robot interactive approach, device and equipment |
CN109656529A (en) * | 2018-10-31 | 2019-04-19 | 北京大学 | A kind of on-line customization method and system for client deep learning |
CN110378463A (en) * | 2019-07-15 | 2019-10-25 | 北京智能工场科技有限公司 | A kind of artificial intelligence model standardized training platform and automated system |
Non-Patent Citations (5)
Title |
---|
KWANG-HYUN PARK等: "A Steward Robot for Human-Friendly Human-Machine Interaction in a Smart House Environment", 《IEEE TRANSACTIONS ON AUTOMATION SCIENCE AND ENGINEERING》 * |
SIAU K等: "Building trust in artificial intelligence, machine learning, and robotics" * |
Y.SEIJI等: "Training AIBO like a Dog - preliminary results", 《RO-MAN 2004. 13TH IEEE INTERNATIONAL WORKSHOP ON ROBOT AND HUMAN INTERACTIVE COMMUNICATION》 * |
王功利: "基于HoloLens的虚拟交互宠物猫的设计与研究" * |
苏丽娟: "基于迁移学习的脑机融合系统的研究", 《中国优秀博硕士学位论文全文数据库(博士)医药卫生科技辑》 * |
Cited By (10)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111461694A (en) * | 2020-03-17 | 2020-07-28 | 上海大学 | Dish identification and pricing system and method based on multi-level learning model |
CN111461694B (en) * | 2020-03-17 | 2023-07-18 | 上海大学 | Dish identification and pricing system and method based on multi-level learning model |
WO2022021421A1 (en) * | 2020-07-31 | 2022-02-03 | Oppo广东移动通信有限公司 | Model management method, system and apparatus, communication device, and storage medium |
CN112214748A (en) * | 2020-10-30 | 2021-01-12 | 支付宝(杭州)信息技术有限公司 | Identity recognition system, method and device |
CN112668687A (en) * | 2020-12-01 | 2021-04-16 | 达闼机器人有限公司 | Cloud robot system, cloud server, robot control module and robot |
WO2022116716A1 (en) * | 2020-12-01 | 2022-06-09 | 达闼机器人股份有限公司 | Cloud robot system, cloud server, robot control module, and robot |
CN112668687B (en) * | 2020-12-01 | 2022-08-26 | 达闼机器人股份有限公司 | Cloud robot system, cloud server, robot control module and robot |
CN112494956A (en) * | 2020-12-15 | 2021-03-16 | 四川长虹电器股份有限公司 | Simulation method and simulation system for converting articles into pets |
CN113435126A (en) * | 2021-07-07 | 2021-09-24 | 魏天骐 | Knowledge sharing processing method, intelligent robot equipment, knowledge sharing system and task learning system |
CN113435126B (en) * | 2021-07-07 | 2024-02-02 | 魏天骐 | Knowledge sharing processing method, intelligent robot device, knowledge sharing system and task learning system |
Also Published As
Publication number | Publication date |
---|---|
CN110866588B (en) | 2023-05-02 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN110866588B (en) | Training learning method and system for realizing individuation of learning ability model of intelligent virtual digital animal | |
Minerva et al. | Digital twin in the IoT context: A survey on technical features, scenarios, and architectural models | |
US8793205B1 (en) | Robotic learning and evolution apparatus | |
Lv et al. | Generative adversarial networks for parallel transportation systems | |
JP7264376B2 (en) | How to generate a general-purpose trained model | |
Acheson et al. | Model based systems engineering for system of systems using agent-based modeling | |
CN111973996B (en) | Game resource release method and device | |
CN108028890A (en) | Crowdsourcing photography is managed in the wireless network | |
CN109145759A (en) | Vehicle attribute recognition methods, device, server and storage medium | |
JP2001222316A (en) | System and method for managing robot | |
CN111368789B (en) | Image recognition method, device, computer equipment and storage medium | |
CN107977625A (en) | A kind of intelligent movable equipment looked for something and intelligent method of looking for something | |
CN101917392A (en) | The communication hub of video game development systems | |
CN102708431A (en) | Game dossier for an advisory services network | |
CN113633983B (en) | Virtual character expression control method and device, electronic equipment and medium | |
CN110414376A (en) | Update method, face recognition cameras and the server of human face recognition model | |
CN107066322B (en) | A kind of online task allocating method towards self-organizing intelligent perception system | |
WO2024067884A1 (en) | Data processing method and related apparatus | |
Gardelli et al. | Designing self-organising environments with agents and artefacts: a simulation-driven approach | |
Terziyan et al. | Causality-aware convolutional neural networks for advanced image classification and generation | |
Zhou | Deep learning‐driven distributed communication systems for cluster online educational platform considering human–computer interaction | |
Ramachandruni et al. | Attentive task-net: Self supervised task-attention network for imitation learning using video demonstration | |
CN110533688A (en) | Follow-on method for tracking target, device and computer readable storage medium | |
CN112862108B (en) | Modularized reinforcement learning model processing method, system, equipment and storage medium | |
CN112257874A (en) | Machine learning method, device and system of distributed machine learning system |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |