WO2018170668A1 - 一种电子虚拟人应用系统的构建方法 - Google Patents

一种电子虚拟人应用系统的构建方法 Download PDF

Info

Publication number
WO2018170668A1
WO2018170668A1 PCT/CN2017/077269 CN2017077269W WO2018170668A1 WO 2018170668 A1 WO2018170668 A1 WO 2018170668A1 CN 2017077269 W CN2017077269 W CN 2017077269W WO 2018170668 A1 WO2018170668 A1 WO 2018170668A1
Authority
WO
WIPO (PCT)
Prior art keywords
data
user
application software
information
client application
Prior art date
Application number
PCT/CN2017/077269
Other languages
English (en)
French (fr)
Inventor
何基强
潘咏春
吴柏建
刘集斌
Original Assignee
信利半导体有限公司
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by 信利半导体有限公司 filed Critical 信利半导体有限公司
Priority to PCT/CN2017/077269 priority Critical patent/WO2018170668A1/zh
Publication of WO2018170668A1 publication Critical patent/WO2018170668A1/zh

Links

Definitions

  • the present invention relates to the field of artificial intelligence, and more particularly to a method of constructing an electronic virtual human application system.
  • the present invention provides a method of constructing an electronic virtual human application system, the application system using a virtual person may be constructed by this method does not effectively solve the problem unaccompanied.
  • a method for constructing an electronic virtual human application system includes the following steps:
  • the client application software sends information to the core model server, and the core model server establishes a generation model framework according to the received information, and feeds back corresponding information to the client after calculation and analysis.
  • Appearance data facial feature data, skin color data, clothing data, facial features data, age data, gender data, sound data, height data, image data, behavioral feature data, voice recognition data, font data, and word library;
  • Data includes head size, neck size, torso size, limb size, height, dwarf, fat, and skinny;
  • the facial features include eyes, nose, mouth, ears, forehead, eyebrows, lips, teeth, and jaw;
  • the skin color includes six different colors, namely red, yellow, brown, blue, black and white; the clothing is classified according to seasonal classification, spring and summer, autumn and winter, and clothing is selected according to different ethnic figures.
  • the Age comprising:
  • Middle age 41-65 years old including: 1) strong period 41-48 years old; 2) stable period 49-55 years old; 3) adjustment period 56-65 years old;
  • Old age After 66 years old, including: 1) old age 67-72 years; 2) middle age 73-84 years old; 3) old age After the age of 85.
  • the various types of information of the virtual character further includes a character category including:
  • a client application software used by the user interacts with the user for voice communication and graphic recognition, plays an entertainment game, and companionship and the soul with the user.
  • client application software receives feedback from the core model server to output sound data, display image data.
  • the core model server is scanned and synthesized through photos.
  • the form of entry is divided into: one or more photos of a photo, a camera image of an instant camera to generate a corresponding file, or a scan by scanning;
  • the client application software copies the previously saved photos and video clips through image processing technology and video cropping, uploads the photos or videos to the core model server for processing, and generates a short video that allows the user to evoke a historic storyline.
  • soul fragments, the sustenance of the soul the client application software recognizes the sound in the audio and video, and the intelligent analysis automatically generates the audio channel of the familiar one of the users, and the user saves the audio and uploads it to the server to retrieve the sound characteristics, and
  • the characters in the sketching are selectively configured; the audio and video channels generated by the client application software are compiled into various logical story scenes according to the user's comprehensive situation through the core model server, and the chat is exchanged. in The user's soul memory is touched, and the user's loved ones are still accompanied.
  • the user sends the required information to the core model server by using the client application software, and the required information also includes weather, address location, encyclopedic knowledge, Chinese meaning, and English knowledge; the user needs the weather.
  • the client application software also reminds the user of the required attention event for the travel of the corresponding weather.
  • the client application software simultaneously reminds the user of the corresponding location information; the client application software integrates English, German, French, Russian , Korean, Japanese, used to facilitate user interaction with different languages in different countries.
  • the core model server rack establishes a database
  • the database includes a comprehensive analysis of the query font database, the voice library, and the word meaning database data
  • the client application software saves the learned and touched events into data.
  • the client application software will query according to the data information provided by the database when it encounters the same scenario, extract effective and valuable data information, and communicate with the user accurately; each virtual character of the user needs
  • the class information is sent to the core model server, and the common features or similarities and differences of a set of data objects in the database are found through classification and clustering, and are classified according to the classification mode to accurately search for data belonging to the same class or classes;
  • the model server establishes the information of the virtual character through the file data packet transmission. In the transmission process, the binary communication number is used as the storage carrier through the set network communication protocol port, and the binary digit analysis is converted into the data that the server can recognize.
  • the client running by the client application software includes: phone, pad, PC, AR, VR.
  • the client application software periodically collects the usage status data of the client, and periodically reminds the user to perform maintenance on the client carried by the client application.
  • the invention establishes a core model server and a client application software for a deceased relative, a missing relative, a distant relative, a Dink family, or a virtual child in the generated family, based on artificial intelligence, and the data amount of the model server will be The iterative superposition will be accumulated continuously.
  • the client application software is installed on the mobile terminal device, and the client application software system installs and registers the electronic information. After the login is completed, the user needs to be customized according to the actual situation of the user, or automatically generated by photos and scans.
  • the data model framework is built according to the relationship, and an interactive data package is generated and uploaded to the remote data analysis model server through the mobile network.
  • the model server receives the data packet, calculates and analyzes the data, and displays the packet status information back to the software mobile terminal.
  • the software mobile terminal After receiving the data, the software mobile terminal separates the complete data packet and parses the data structure one by one, correspondingly filling Each data structure branch in the data model displays the output.
  • the configuration information can be changed and added at any time.
  • the application software system is not restricted by the mobile terminal platform, and the model packet information of the configuration that has been established is loaded at any time, and the connection with the server network is kept unblocked, and the status information is obtained in real time.
  • FIG. 1 is a flowchart of the present method of the invention
  • Figure 2 is a signal trend diagram of the present invention
  • FIG. 3 is a structural diagram of a hardware system of a system built by the present invention.
  • FIG. 4 is a schematic diagram showing the operation of the hardware system of the system constructed by the present invention.
  • a method for constructing an electronic virtual human application system includes the following steps:
  • the client application software sends information to the core model server, and the core model server establishes a generation model framework according to the received information, and feeds back corresponding information to the client after calculation and analysis.
  • the present invention realizes data exchange by establishing a data link between a client application software system and a core model server in a wired network or a mobile wireless network.
  • the client application software is installed on the mobile software terminal, and the electronic information is registered by connecting the server, and the account information can be established for the user, and is not limited to the hardware terminal device platform, and can be easily carried, installed, registered, and logged in. Load and obtain user configuration information, and save the original established packet model information in local storage.
  • the method of the present invention is not limited to the use of hardware device platforms, such as phone, pad, PC, AR, VR type devices.
  • the user can customize the configuration information according to the actual situation of the user, and the settings include the following:
  • Appearance, facial features, skin color, clothing, gender, age, sound characteristics, speech rate, height, fat and thin, date of birth in addition, can also automatically generate preliminary character morphological features through photo entry and scanning.
  • the user can establish a body shape and a face feature according to the character setting.
  • the user can configure the style and style of the costume according to the character.
  • the clothing is seasonal, and in addition, clothing can be selected according to different ethnic groups.
  • Gender age stage
  • height height
  • speed of speech rate and sound characteristics are arbitrarily configured according to user needs.
  • the electronic virtual human application software system may choose to automatically generate data through a photo, scan form, upload to a data model server, receive a data packet, and establish a data model framework,
  • the data model server processes the analysis calculations and generates a data packet.
  • the client application software establishes a data communication mechanism with the remote server through the network to save the data to the local storage.
  • the data packet calculated and analyzed is obtained locally, and after parsing, the output is displayed.
  • the client application software sends the data to the model server and through data parsing, after the initialization of the data is completed, the user performs human-computer interaction with the electronic virtual person.
  • the electronic device terminal receives the user feeding the electronic virtual person application software system; the voice communication receiving, the graphic recognition interaction, the playing entertainment game, the companionship and the soul comforting with the user, the medicine reminding, the smoking cessation advice, the chat function, etc. Certain events.
  • the model server inputs these data, establishes an electronic virtual human model, and obtains the next state and excessive state data of the electronic virtual person.
  • the electronic virtual human model server calculates the electronic virtual human state data through the analysis system according to the received data, and finally displays the corresponding state data, including outputting sounds, images, and the like.
  • the client application software receives the parsed state data, and combines the state data information to output sound data and image data accordingly.
  • the information processing analysis system and information storage are supported by artificial intelligence services.
  • the information analysis processing system analyzes and stores data information through human sexual behavior and social psychology artificial intelligence, and performs comparison, extraction, clustering, and optimization. After system analysis, provide users with information functions that are more realistic, effective, personalized and proactive.
  • the analysis processing system is a combination of behavioral and psychology, and includes an active mode and a passive mode service, which can meet the needs of users personalized in the current information application.
  • the user usually likes to read and watch some of the latest domestic and foreign information news every day for a certain period of time, according to the user's usual preference, record in the model server, search through the data analysis system, actively exchange information services to the user, and the user The required information is displayed in the window.
  • the model server transmits the status data to the service terminal installed with the electronic virtual human application software in real time, and the electronic virtual human model server transmits the obtained next state and the transition intermediate data information to the service terminal through the network data transmission protocol.
  • Each end user is an independently owned personalized requirement setting with its own specificity and characteristics.
  • Each terminal has one or more independent personalized electronic virtual person communication platforms.
  • the communication platform is independent, friendly, self-contained, easy to use, easy to interact with humans, and capable of self-growth and self-learning functions to meet the individual needs of users.
  • the client application software terminal can scan and synthesize the input system through photos, and automatically match and generate.
  • the client application software terminal uploads a photo or video to the model server for processing, and generates a short video.
  • the client application software terminal recognizes the sound in the audio and video, intelligently analyzes the sound characteristics, generates the audio channel of the familiar person of the user, saves the audio, and uploads it to the server, and the character can select to configure the audio as the voice to start. Channel.
  • the client application software terminal compiles the generated audio and video channels into various scenes with logical stories according to the user's comprehensive situation.
  • the client application software terminal prompts the user in advance during the birthday, and lets the user celebrate the electronic virtual person.
  • the client application software terminal can request query assistance, weather inquiry, address location query, encyclopedic knowledge query, Chinese word query, and English knowledge query to the electronic virtual person by issuing a voice instruction.
  • the client application software terminal, the system can not only help the query, but also perform human-computer interaction and chat conversation with the user according to different query scenarios.
  • the electronic virtual person will care for the user according to the weather conditions, and remind the user to pay attention to the event;
  • Such as 2 The address location query, the electronic virtual person through the user to query the address information, friendly communication with the user whether to choose an address for a certain day, prompting the user to pay attention to some events or help the user to query more people, texts and other related information of this address.
  • the client application software terminal can arbitrarily select a language for different users, and communicate with the electronic virtual person software application system, which runs through the entire application software system.
  • the client application software terminal and the user communicate each other by voice to send a response, telling the user that the electronic virtual person is hungry to feed, and time to take medicine.
  • the client application software terminal communicates emotionally, and finally gets married, the electronic virtual person actively sends a request instruction, and the server receives the request and matches the marriage.
  • the model server After the client application software terminal plays with the friend for a period of time, the model server sends a status message indicating that 'dirty' is displayed according to the analysis according to the shape, and reminds the user to clean and bathe the electronic virtual person.
  • the client application software terminal distinguishes between adult and underage, and has different care and care for the electronic virtual person, and the user will remind him of the operation of getting up, washing, and feeding.
  • the client application software terminal has a self-learning function.
  • the user can recommend books, audio, video, etc. for the electronic virtual person, so that the electronic virtual person can self-read and self-learn. And can absorb the knowledge information that is learned, so that more exchanges and exchanges of information with each other; the various types of information of the virtual characters that the user needs are sent to the core model server, and the database is found by classification and clustering.
  • the binary number is used as the storage carrier, and the binary digit analysis is converted into the data that the server can recognize.
  • the present invention provides solutions and implementations for deceased relatives, missing relatives, distant relatives, Dink family, or virtual children in the family. Based on artificial intelligence, the amount of data information of the model server is continuously accumulated and iteratively superimposed. Relying on hardware, software and hardware are combined to realize, electronic virtual human application software system implementation and method, and the implementation method can be realized by using the existing technology reasonably and conveniently.
  • the present invention can be ideally served in an active and passive mode, and can satisfy the user's personalized needs in the current information application, and achieve an ideal mode for human-machine exchange in the future, achieving human-machine friendly communication.

Abstract

本发明提供一种电子虚拟人应用系统的构建方法,该方法为逝去的亲人、失踪的亲人、远离的亲人、丁克家族、或生成家庭中虚拟的孩子等,建立核心模型服务器和客户端应用软件,以人工智能为基础,模型服务器的数据信息量将会不断积累迭代叠加,客户端应用软件安装于移动终端设备,客户端应用软件系统安装注册电子信息,登陆完成后,能自定义根据用户实际情况配置人物需求信息,或通过照片、扫描自动生成,根据关系建立数据模型框架,生成一个交互数据包,并通过移动网络上传到远程数据分析模型服务器。

Description

一种电子虚拟人应用系统的构建方法
技术领域
[0001] 本发明涉及人工智能领域,更具体地,涉及一种电子虚拟人应用系统的构建方法。
背景技术
[0002] 随着互联网技术高速发展,移动互联网数据增值业务从 2G 、 3G 、 4G 等网络不断往前推进。手机、消费电子产品在现有基础设施下,跨入新时代,逐渐融入到百姓们的生活,设备终端、软件终端功能以及增值业务的日益增多,至今形成对智能产品,应用软件系统的信任与依赖,人与人之间的沟通方式有了创新的改变,人与机器建立相互通信的机制管道,电子虚拟人应用系统在此背景下,实现移动互联,数据互通,数据互传,实时获取远端服务器状态数据信息,与达到电子虚拟人智能应用软件系统友好的与人沟通,使得用户获得了愉悦的操作享受。
[0003] 为逝去的亲人、失踪的亲人、远离的亲人、丁克家族解决亲人陪伴、伴护的问题日益凸显,如何实现根据用户需求随意增加或更换电子虚拟人应用软件,融入到用户的生活,这将是成为待解决的问题。
发明内容
[0004] 本发明提供一种电子虚拟人应用系统的构建方法,利用该方法构建的虚拟人应用系统可有效解决没有亲人陪伴的问题。
[0005] 为了达到上述技术效果,本发明的技术方案如下:
一种电子虚拟人应用系统的构建方法,包括以下步骤:
S1 :建立以人工智能为基础核心模型服务器,该核心模型服务器收集、分析和处理要求建立的虚拟人物的信息;
S2 :搭建用户使用的客户端应用软件,用于与核心模型服务器交互,向核心模型服务器发送用户自身需要的虚拟人物的各类信息;
S3 :客户端应用软件向核心模型服务器发送信息,核心模型服务器根据接收到信息建立生成模型框架,经计算分析处理后向客户端反馈相应的信息。
[0006] 进一步地,所述用户自身需要的虚拟人物的各类信息包括:
外观数据、面貌特征数据、肤色数据、服装数据、五官数据、年龄数据、性别数据、声音数据、身高数据、图像数据、行为特征数据、语音识别数据、字库数据和词语字意库;所述外观数据包括头部尺寸、颈部尺寸、躯干尺寸、四肢尺寸、高、矮、胖、瘦;所述面貌特征包括眼睛、鼻子、嘴巴、耳朵、额头、眼眉、嘴唇、牙齿和下颏;所述肤色包括六种不同的颜色,即红、黄、棕、蓝、黑和白色;所述服装根据季节性分类春夏装、秋冬装,还根据不同种族人物选择设置服装。
[0007] 进一步地,所述年龄包括:
童年 :0 岁 -6 岁包括: 1) 婴儿期 0-3 周月; 2) 小儿期 4 周月 -2.5 岁; 3) 幼儿期 2.5 岁后 -6 岁;
少年 :7 岁- 17 岁包括: 1) 启蒙期 7 岁 -10 岁; 2) 逆反期 11 岁 -14 岁; 3) 成长期 15 岁 -17 岁;
青年 :18 岁 -40 岁包括: 1) 青春期 18-28 岁; 2) 成熟期 29-40 岁;
中年 :41-65 岁包括: 1) 壮实期 41-48 岁; 2) 稳健期 49-55 岁; 3) 调整期 56-65 岁;
老年 :66 岁以后包括: 1) 初老期 67-72 岁; 2) 中老期 73-84 岁; 3) 年老期 85 岁以后。
[0008] 进一步地,虚拟人物的各类信息还包括人物类别包括:
逝去的亲人、失踪的亲人、远离的亲人、丁克家族、或生成家庭中虚拟的孩子。
[0009] 进一步地,利用人性行为学和社会心理学搭建用户使用的客户端应用软件,用户使用的客户端应用软件与用户进行语音交流和图形识别交互,玩耍娱乐游戏,与用户进行陪伴及心灵藉慰,吃药提醒,戒烟劝诫,聊天;客户端应用软件接收核心模型服务器反馈的回的信息输出声音数据、显示图像数据。
[0010] 进一步地,根据一些丁克家族、未生育、未能生育人群用户,通过照片,扫描合成录入核心模型服务器。
[0011] 其录入形式分为:照片一张或多张合成、即时摄像头拍照录像生成相应的文件、或通过扫描录入;
客户端应用软件将以往留存的照片、视频短片,通过图像处理技术、视频裁剪,将照片或视频上传至核心模型服务器加工处理后,生成一段小短片,能让用户勾起有历史性的故事情节,灵魂碎片,心灵的寄托;客户端应用软件将音频、视频中的声音识别,智能分析自动生成用户所熟悉亲人的音频声道,用户将音频保存,上传至服务器,以便调取声音特性,与勾画中的人物有可选择性的配置;客户端应用软件将生成的音频、视频的声道,经核心模型服务器根据用户综合情况,编成各种带有逻辑性的故事的场景,在聊天交流中 , 触感用户的灵魂记忆,感觉到用户的亲人依然陪伴着。
[0012] 进一步地,用户利用客户端应用软件通过语音将需要的信息发送至核心模型服务器,所述需要的信息的还包括天气、地址位置、百科知识、汉语字意、英文知识;用户需要天气信息时,客户端应用软件同时提醒用户相应天气的出行所需注意事件,户需要地址位置信息时,客户端应用软件同时提醒用户相应位置的信息;客户端应用软件集成英语、德语、法语、俄语、韩语、日语,用于方便与不同国家不同语言的用户交互。
[0013] 进一步地,所述核心模型服务器架建立数据库,数据库包括查询字库、语音库、字词字意库数据的综合分析,客户端应用软件将所学习过、接触过的事件,保存成数据链条,客户端应用软件在遇到同样的场景时,会根据数据库所能提供的数据信息进行查询,提取有效,有价值的数据信息,与用户准确的交流沟通;用户自身需要的虚拟人物的各类信息向核心模型服务器发送 , 通过分类、聚类找出数据库中一组数据对象的共同特点或相似性和差异性,并按照分类模式将其划分,准确寻找属同一类或几类的数据;模型服务器建立虚拟人物的信息通过文件数据包的形式传输,在传输过程中通过设定的网络通信协议端口,以二进制数字作为存储载体,并将二进制数字解析转成服务器能识别的数据。
[0014] 优选地,客户端应用软件所运行的客户端包括: phone 、 pad 、 PC 、 AR , VR 。
[0015] 进一步地,用客户端应用软件定期收集本客户端的使用状况数据,并定期提醒用户对该客户端应用软件所搭载的客户端进行维护。
[0016] 与现有技术相比,本发明技术方案的有益效果是:
本发明为逝去的亲人、失踪的亲人、远离的亲人、丁克家族、或生成家庭中虚拟的孩子等,建立核心模型服务器和客户端应用软件,以人工智能为基础,模型服务器的数据信息量将会不断积累迭代叠加,客户端应用软件安装于移动终端设备,客户端应用软件系统安装注册电子信息,登陆完成后,能自定义根据用户实际情况配置人物需求信息,或通过照片、扫描自动生成,根据关系建立数据模型框架,生成一个交互数据包,并通过移动网络上传到远程数据分析模型服务器。经模型服务器接收到数据包,计算分析,将数据显示包状态信息,回传到软件移动终端,软件移动终端接收到数据后,将完整的数据包,进行逐个数据结构拆开解析,对应地填充数据模型中的每个数据结构分支,并显示输出。配置信息可以随时更改、增加,应用软件系统不受移动终端平台限制,随时登陆加载配置已建立完成的模型数据包信息,保持与服务器网络连接畅通,实时获取状态信息。
附图说明
[0017] 图 1 为本发明方法流程图;
图 2 为本发明中信号走向图;
图 3 为本发明搭建的系统的硬件系统结构图;
图 4 为本发明搭建的系统的硬件系统的工作示意图。
具体实施方式
[0018] 附图仅用于示例性说明,不能理解为对本专利的限制;
为了更好说明本实施例,附图某些部件会有省略、放大或缩小,并不代表实际产品的尺寸;
对于本领域技术人员来说,附图中某些公知结构及其说明可能省略是可以理解的。
[0019] 下面结合附图和实施例对本发明的技术方案做进一步的说明。
[0020] 实施例 1
如图 1-2 所示,一种电子虚拟人应用系统的构建方法,包括以下步骤:
S1 :建立以人工智能为基础核心模型服务器,该核心模型服务器收集、分析和处理要求建立的虚拟人物的信息;
S2 :搭建用户使用的客户端应用软件,用于与核心模型服务器交互,向核心模型服务器发送用户自身需要的虚拟人物的各类信息;
S3 :客户端应用软件向核心模型服务器发送信息,核心模型服务器根据接收到信息建立生成模型框架,经计算分析处理后向客户端反馈相应的信息。
[0021] 本发明根据用户的需求随意增加、更换终端上的客户端应用软件,丰富了数据增值业务,实现移动互联,数据互通,数据互传。
[0022] 如图 3-4 所示,本发明通过在有线网络、移动无线网络中,客户端应用软件系统与核心模型服务器建立数据链路,实现数据交换。
[0023] 将客户端应用软件安装于移动软件终端,通过连接服务器实现电子信息注册,并能为用户建立账号信息,不受限于硬件终端设备平台,就能够便于携带,安装、注册、登陆,加载获取用户配置信息,将原先建立好的数据包模型信息保存于本地存储器。
[0024] 本发明方法不受限于硬件设备平台,比如, phone 、 pad 、 PC 、 AR 、 VR 类设备的使用运行。
[0025] 用户完成登录后,用户能自定义根据用户实际情况配置需求信息,设置包括以下:
外观,面貌特征,肤色,服装,性别,年龄,声音特征,语速,身高,胖瘦,出生日期,除此外,还可以通过照片录入、扫描自动生成初步人物形态特征等。
[0026] 用户可根据人物设置建立外观体型,面貌特征。
[0027] 用户可根据人物设置服装样式、风格随意配置。分类服装季节性,此外,还能根据不同种族人物选择设置服装。人物照片服装款式、风格扫描录入系统,智能化、个性化自动匹配。
[0028] 性别、年龄阶段、身高、语速快慢节奏、声音性质特征视用户需求任意配置。
[0029] 电子虚拟人应用软件系统,为不能完全勾画出人物对象信息的用户,可以选择通过照片、扫描的形式自动生成数据,上传至数据模型服务器,接收到数据包,建立数据模型框架,由数据模型服务器处理分析计算,生成一个数据包。
[0030] 客户端应用软件通过网络,与远端服务器建立数据通信机制,将数据保存至本地存储器。
[0031] 经计算分析的数据包获取到本地,解析后,显示输出。
[0032] 运行中的电子虚拟人,由用户发出指令,经数据交换计算出当前状态和下一个状态,将数据实时更新在需要显示输出的终端上。
[0033] 客户端应用软件将数据发送至模型服务器并通过数据解析,完成数据的初始化之后,用户便与电子虚拟人进行人机交互。例如饲养,电子设备终端接收用户给电子虚拟人应用软件系统喂食物事件;语音交流接收、图形识别交互,玩耍娱乐游戏,与用户进行陪伴及心灵藉慰,吃药提醒,戒烟劝诫,聊天功能等某些事件。
[0034] 模型服务器把这些数据输入,建立电子虚拟人模型,得到电子虚拟人的下一个状态和过度状态数据。电子虚拟人模型服务器根据收到的数据,经过分析系统的计算电子虚拟人状态数据,并最终展示相应的状态数据,包括输出声音、图像等。
[0035] 客户端应用软件收到解析后的状态数据,结合状态数据信息,相应的输出声音数据、图像数据。
[0036] 终端用户建立或选定自己喜爱的人物。
[0037] 建立以人性行为学,社会心理学作为基础的信息处理分析系统。
[0038] 以信息处理分析系统和信息存储,人工智能服务作为依托。
[0039] 信息分析处理系统,通过人性行为学,社会心理学人工智能分析存储数据信息,进行对比,抽取、聚类,优化。在系统分析后,面向用户提供更加真实、有效、个性化和主动性服务的信息功能。
[0040] 所述分析处理系统为行为学和心理学结合,包含主动式模式和被动式模式服务,能满足用户在当前信息化应用中个性化的需求。
[0041] 比如:用户平时每天在某时间段比较喜欢阅读观看一些最新国内外资讯新闻,根据用户平时喜好,记录于模型服务器,经数据分析系统查找,主动式向用户交换信息服务,将用户所需要的信息资料,推送显示在窗口当中。
模型服务器把这些状态数据实时传送给安装有电子虚拟人应用软件的服务终端,电子虚拟人模型服务器通过网络数据传输协议,把得到的下一个状态以及过渡的中间数据信息传送给服务终端。
[0042] 每个终端用户属独立拥有的个性化需求化设置,有自己专属性和特征性。每个终端拥有一个或多个独立个性化电子虚拟人交流平台。
[0043] 使得这个交流平台具有独立、亲和、自主发挥、易用、便于人机交互,并能够自我成长,自我学习功能,满足用户个性化需求。
[0044] 客户端应用软件终端,根据一些丁克家族、未生育、未能生育等人群用户,可以通过照片,扫描合成录入系统,自动匹配生成。
[0045] 客户端应用软件终端,将照片或视频上传至模型服务器加工处理后,生成一段小短片。
[0046] 客户端应用软件终端,将音频、视频中的声音识别,智能分析,声音特性,生成用户所熟悉亲人的音频声道,音频保存,上传至服务器,人物可以选择配置音频作为语音出发的声道。
[0047] 客户端应用软件终端,将生成的音频、视频的声道,经模型服务器根据用户综合情况,编成各种带有逻辑性的故事的场景。
[0048] 客户端应用软件终端,到了生日期间提前提示用户,让用户为电子虚拟人庆贺。
[0049] 客户端应用软件终端,可以通过发出语音指令,向电子虚拟人请求查询帮助,天气查询、地址位置查询、百科知识查询、汉语字意查询、英文知识查询等。
[0050] 客户端应用软件终端,该系统不但可以帮助查询,还能根据不同的查询情景跟用户进行人机交互,聊天对话。
[0051] 如 1 ,天气查询,电子虚拟人会根据天气情况,关怀用户,提醒用户出行所需注意事件;
如 2 ,地址位置查询,电子虚拟人通过用户查询地址信息,友好的与用户交流是否某天选择出行某地址,提示用户该注意些事件或帮助用户查询更多此地址的人、文等相关信息。
[0052] 客户端应用软件终端,针对不同用户可任意选择语言,与电子虚拟人软件应用系统进行交流,所述贯通整个应用软件系统。
[0053] 客户端应用软件终端与用户之间通过声音传达相互呼唤回应,告诉用户电子虚拟人肚子饿要喂食、时间到要吃药。
[0054] 搜索附近周边的电子虚拟人,之间建立好友关系。
[0055] 电子虚拟人与好友之间的关系,进行玩耍、结婚、生子、领养、聊天、社会交际。
[0056] 不同的两个客户端应用软件终端趣味交互和情感交流,发出指令,通过模型服务器进行匹配,生产一个新生儿,电子虚拟人。
[0057] 客户端应用软件终端间情感交流,最终结婚,电子虚拟人主动发出请求指令 , 服务器接收请求,匹配结婚。
[0058] 客户端应用软件终端与好友玩耍一段时间后,模型服务器经根据状分析,发出显示'脏'的状态信息,提醒用户要为电子虚拟人清洁、洗澡。
[0059] 客户端应用软件终端,区分成年与未成年,对电子虚拟人有不同的照顾方式和关爱,用户将为它提醒起床、洗漱、喂食的操作。
[0060] 客户端应用软件终端,还有自我学习功能,根据电子虚拟人所的掌握信息量,用户可为电子虚拟人推荐书籍、音频、视频等,让电子虚拟人的自我阅读、自我学习能力,并能将所学习的知识信息吸收了,以便更多的与用相互户交流、交换信息;用户自身需要的虚拟人物的各类信息向核心模型服务器发送 , 通过分类、聚类找出数据库中一组数据对象的共同特点或相似性和差异性,并按照分类模式将其划分,准确寻找属同一类或几类的数据;模型服务器建立虚拟人物的信息通过文件数据包的形式传输,在传输过程中通过设定的网络通信协议端口,以二进制数字作为存储载体,并将二进制数字解析转成服务器能识别的数据。
[0061] 本发明为逝去的亲人、失踪的亲人、远离的亲人、丁克家族、或生成家庭中虚拟的孩子等,提供解决实现及方法。以人工智能为基础,模型服务器的数据信息量 , 不断积累迭代叠加。以硬件作为依托,软件和硬件相互结合来实现,电子虚拟人应用软件系统实现及方法,实现方式可利用现已有的技术合理方便的实现。
[0062] 本发明能以主动式和被动式模式理想服务,能够满足用户在当前信息化应用中有个性化的需求,为今后人机交换理想模式,达到人与机器友好的沟通。
[0063] 相同或相似的标号对应相同或相似的部件;
附图中描述位置关系的用于仅用于示例性说明,不能理解为对本专利的限制;
显然,本发明的上述实施例仅仅是为清楚地说明本发明所作的举例,而并非是对本发明的实施方式的限定。对于所属领域的普通技术人员来说,在上述说明的基础上还可以做出其它不同形式的变化或变动。这里无需也无法对所有的实施方式予以穷举。凡在本发明的精神和原则之内所作的任何修改、等同替换和改进等,均应包含在本发明权利要求的保护范围之内。

Claims (10)

  1. 一种电子虚拟人应用系统的构建方法,其特征在于,包括以下步骤:
    S1 :建立以人工智能为基础核心模型服务器,该核心模型服务器收集、分析和处理要求建立的虚拟人物的信息;
    S2 :搭建用户使用的客户端应用软件,用于与核心模型服务器交互,向核心模型服务器发送用户自身需要的虚拟人物的各类信息;
    S3 :客户端应用软件向核心模型服务器发送信息,核心模型服务器根据接收到信息建立生成模型框架,经计算分析处理后向客户端反馈相应的信息。
  2. 根据权利要求 1 所述的电子虚拟人应用系统的构建方法,其特征在于,所述用户自身需要的虚拟人物的各类信息包括: 外观数据、面貌特征数据、 肤色数据、服装数据、 五官数据、年龄数据、性别数据、声音数据、身高数据、图像数据、行为特征数据、语音识别数据、字库数据和词语字意库;所述外观数据包括 头部尺寸、颈部尺寸、躯干尺寸、四肢尺寸、高、矮、胖、瘦;所述面貌特征包括眼睛、鼻子、嘴巴、耳朵、额头、眼眉、嘴唇、牙齿和下颏;所述肤色包括六种不同的颜色,即红、黄、棕、蓝、黑和白色;所述服装根据季节性分类春夏装、秋冬装,还根据不同种族人物选择设置服装。
  3. 根据权利要求 2 所述的电子虚拟人应用系统的构建方法,其特征在于,所述年龄包括:
    童年 :0 岁 -6 岁包括: 1) 婴儿期 0-3 周月; 2) 小儿期 4 周月 -2.5 岁; 3) 幼儿期 2.5 岁后 -6 岁;
    少年 :7 岁 -17 岁包括: 1) 启蒙期 7 岁 -10 岁; 2) 逆反期 11 岁 -14 岁; 3) 成长期 15 岁 -17 岁;
    青年 :18 岁 -40 岁包括: 1) 青春期 18-28 岁; 2) 成熟期 29-40 岁;
    中年 :41-65 岁包括: 1) 壮实期 41-48 岁; 2) 稳健期 49-55 岁; 3) 调整期 56-65 岁;
    老年 :66 岁以后包括: 1) 初老期 67-72 岁; 2) 中老期 73-84 岁; 3) 年老期 85 岁以后。
  4. 根据权利要求 3 所述的电子虚拟人应用系统的构建方法,其特征在于,虚拟人物的各类信息还包括 人物类别包括: 逝去的亲人、失踪的亲人、远离的亲人、丁克家族、或生成家庭中虚拟的孩子。
  5. 根据权利要求 4 所述的电子虚拟人应用系统的构建方法,其特征在于,利用 人性行为学和社会心理学搭建 用户使用的客户端应用软件 , 用户使用的客户端应用软件与用户进行 语音交流和图形识别交互,玩耍娱乐游戏,与用户进行陪伴及心灵藉慰,吃药提醒,戒烟劝诫,聊天; 客户端应用软件接收核心模型服务器反馈的回的信息 输出声音数据、显示图像数据。
  6. 根据权利要求 5 所述的电子虚拟人应用系统的构建方法,其特征在于, 根据一些丁克家族、未生育、未能生育人群用户,通过照片,扫描合成录入 核心模型服务器 ;
    其录入形式分为:照片一张或多张合成、即时摄像头拍照录像生成相应的文件、或通过扫描录入;
    客户端应用软件 将以往留存的照片、视频短片,通过图像处理技术、视频裁剪,将照片或视频上传至 核心模型服务器 加工处理后,生成一段小短片,能让用户勾起有历史性的故事情节,灵魂碎片,心灵的寄托; 客户端应用软件 将音频、视频中的声音识别,智能分析自动生成用户所熟悉亲人的音频声道,用户将音频保存,上传至服务器,以便调取声音特性,与勾画中的人物有可选择性的配置; 客户端应用软件 将生成的音频、视频的声道,经 核心模型服务器 根据用户综合情况,编成各种带有逻辑性的故事的场景,在聊天交流中,触感用户的灵魂记忆,感觉到用户的亲人依然陪伴着。
  7. 根据权利要求6所述的电子虚拟人应用系统的构建方法,其特征在于,用户利用客户端应用软件通过语音将需要的信息发送至核心模型服务器,所述需要的信息的还包括天气、地址位置、百科知识、汉语字意、英文知识;用户需要天气信息时,客户端应用软件同时提醒用户相应天气的出行所需注意事件,户需要地址位置信息时,客户端应用软件同时提醒用户相应位置的信息;客户端应用软件集成英语、德语、法语、俄语、韩语、日语,用于方便与不同国家不同语言的用户交互。
  8. 根据权利要求7所述的电子虚拟人应用系统的构建方法,其特征在于,所述核心模型服务器架建立数据库,数据库包括查询字库、语音库、字词字意库数据的综合分析,客户端应用软件将所学习过、接触过的事件,保存成数据链条,客户端应用软件在遇到 同样的场景时,会根据数据库所能提供的数据信息进行查询,提取有效,有价值的数据信息,与用户准确的交流沟通; 用户自身需要的虚拟人物的各类信息向核心模型服务器发送 , 通过分类、聚类找出数据库中一组数据对象的共同特点或相似性和差异性,并按照分类模式将其划分,准确寻找属同一类或几类的数据;模型服务器建立虚拟人物的信息通过文件数据包的形式传输,在传输过程中通过设定的网络通信协议端口,以二进制数字作为存储载体,并将二进制数字解析转成服务器能识别的数据。
  9. 根据权利要求8所述的电子虚拟人应用系统的构建方法,其特征在于,客户端应用软件所运行的客户端包括: phone 、pad、PC、AR,VR。
  10. 根据权利要求9所述的电子虚拟人应用系统的构建方法,其特征在于,用客户端应用软件定期收集本客户端的使用状况数据,并定期提醒用户对该客户端应用软件所搭载的客户端进行维护。
PCT/CN2017/077269 2017-03-20 2017-03-20 一种电子虚拟人应用系统的构建方法 WO2018170668A1 (zh)

Priority Applications (1)

Application Number Priority Date Filing Date Title
PCT/CN2017/077269 WO2018170668A1 (zh) 2017-03-20 2017-03-20 一种电子虚拟人应用系统的构建方法

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
PCT/CN2017/077269 WO2018170668A1 (zh) 2017-03-20 2017-03-20 一种电子虚拟人应用系统的构建方法

Publications (1)

Publication Number Publication Date
WO2018170668A1 true WO2018170668A1 (zh) 2018-09-27

Family

ID=63584591

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2017/077269 WO2018170668A1 (zh) 2017-03-20 2017-03-20 一种电子虚拟人应用系统的构建方法

Country Status (1)

Country Link
WO (1) WO2018170668A1 (zh)

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111290570A (zh) * 2018-12-10 2020-06-16 中国移动通信集团终端有限公司 人工智能设备的控制方法、装置、设备及介质
CN111496802A (zh) * 2019-01-31 2020-08-07 中国移动通信集团终端有限公司 人工智能设备的控制方法、装置、设备及介质
CN111984277A (zh) * 2020-08-21 2020-11-24 江苏鑫合易家信息技术有限责任公司 一种统计app安装量的方法

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103218844A (zh) * 2013-04-03 2013-07-24 腾讯科技(深圳)有限公司 虚拟形象的配置方法、实现方法、客户端、服务器及系统
US20130198210A1 (en) * 2012-01-27 2013-08-01 NHN ARTS Corp. Avatar service system and method provided through a network
CN105260588A (zh) * 2015-10-23 2016-01-20 福建优安米信息科技有限公司 一种健康守护机器人系统及其数据处理方法

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20130198210A1 (en) * 2012-01-27 2013-08-01 NHN ARTS Corp. Avatar service system and method provided through a network
CN103218844A (zh) * 2013-04-03 2013-07-24 腾讯科技(深圳)有限公司 虚拟形象的配置方法、实现方法、客户端、服务器及系统
CN105260588A (zh) * 2015-10-23 2016-01-20 福建优安米信息科技有限公司 一种健康守护机器人系统及其数据处理方法

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111290570A (zh) * 2018-12-10 2020-06-16 中国移动通信集团终端有限公司 人工智能设备的控制方法、装置、设备及介质
CN111496802A (zh) * 2019-01-31 2020-08-07 中国移动通信集团终端有限公司 人工智能设备的控制方法、装置、设备及介质
CN111984277A (zh) * 2020-08-21 2020-11-24 江苏鑫合易家信息技术有限责任公司 一种统计app安装量的方法

Similar Documents

Publication Publication Date Title
WO2020213799A1 (ko) 스마트 휴대 기기를 이용한 스마트 기기와 로봇의 개인 맞춤형 서비스 운용 시스템 및 방법
Scheflen Human communication: Behavioral programs and their integration in interaction
US5982853A (en) Telephone for the deaf and method of using same
US8285552B2 (en) System and method for simulating expression of message
WO2018170668A1 (zh) 一种电子虚拟人应用系统的构建方法
CN208141466U (zh) 一种基于机器人的签到装置及系统
CN109129509A (zh) 一种基于屏幕智能交互的养老陪护机器人
CN106570327B (zh) 一种湿疹宝宝病情管理系统
WO2015184806A1 (zh) 一种健康状况的改善方法、装置及系统
CN107734213A (zh) 智能家用电子装置与系统
CN107168174B (zh) 一种使用机器人做居家养老的方法
WO2016200141A1 (ko) 가족 문화 컨텐츠 서비스 방법 및 시스템
CN109523852A (zh) 基于视觉监控的学习交互系统及其交互方法
Taketomo Amae as metalanguage: A critique of Doi's theory of amae
Taylor et al. Commercialisation and commodification of breastfeeding: video diaries by first-time mothers
CN108140045A (zh) 在增强和替代通信系统中支持感知和对话处理量
CN110178159A (zh) 具有集成式投影仪的音频/视频可穿戴式计算机系统
JP2017119336A (ja) ロボット、ロボット制御方法並びにロボットプログラム
WO2020122463A1 (ko) 가상 현실을 이용한 치매 환자 훈련 시스템
CN114201035A (zh) 信息处理装置、存储介质及信息处理方法
CN112863407A (zh) 智能触摸拼接屏人机交互式系统及方法
JP2003004863A (ja) 在席情報伝達装置および在席情報伝達方法
WO2020116888A1 (ko) 네트워크 기반의 이동형 식당 테이블
JP4004808B2 (ja) 新生児情報電送システム
US20230394878A1 (en) Program, information processing device, and method

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 17901676

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 17901676

Country of ref document: EP

Kind code of ref document: A1