CN115392332A

CN115392332A - AI model deployment method, system and storage medium

Info

Publication number: CN115392332A
Application number: CN202110574310.4A
Authority: CN
Inventors: 潘绪洋; 阳慢情; 李志强; 赵丛
Original assignee: Gongdadi Innovation Technology Shenzhen Co ltd
Current assignee: Gongdadi Innovation Technology Shenzhen Co ltd
Priority date: 2021-05-25
Filing date: 2021-05-25
Publication date: 2022-11-25

Abstract

The embodiment of the invention provides an AI model deployment method, a server, display equipment, a system and a storage medium, belonging to the field of artificial intelligence. The method comprises the following steps: acquiring a training data set and an initial AI model to be trained; training the initial AI model based on a training data set to generate a target AI model; receiving a model issuing instruction sent by display equipment; analyzing the model issuing command to acquire first identification information of target terminal equipment of the model to be deployed and second identification information of the target AI model to be issued; and executing an AI model issuing instruction according to the first identification information and the second identification information so as to allow the target terminal equipment to receive and load the target AI model. The target AI model is quickly and conveniently sent to the target terminal equipment through the AI model issuing page, so that the deployment flexibility and convenience of the AI model are greatly improved, and a user can quickly and conveniently use the AI model on the terminal.

Description

AI model deployment method, system and storage medium

Technical Field

The invention relates to the field of artificial intelligence, in particular to an AI model deployment method, a server, display equipment, a system and a storage medium.

Background

With the rapid development of artificial intelligence, the AI model is widely applied to the fields of face recognition, image classification, object detection, speech recognition, natural language processing, and the like. At present, people mainly burn the AI model into terminal equipment through a data line or other modes, so that the terminal equipment can operate the AI model to perform face recognition, image classification, object detection, voice recognition and the like, but the conventional method for burning the AI model into the terminal equipment is too complicated, flexible updating and replacement of the AI model in the terminal equipment are not facilitated, and the user experience is not good.

Disclosure of Invention

The embodiment of the invention provides an AI model deployment method, a server, a display device, a system and a storage medium, which are beneficial to improving the deployment flexibility of an AI model in terminal equipment.

In a first aspect, an embodiment of the present invention provides an AI model deployment method, which is applied to a server, and includes:

acquiring a training data set and an initial AI model to be trained;

training the initial AI model based on the training data set to generate a target AI model;

receiving a model issuing instruction sent by display equipment, wherein the model issuing instruction is generated by the display equipment according to the operation of a user on a displayed AI model issuing page;

analyzing the model issuing command to acquire first identification information of target terminal equipment of a model to be deployed and second identification information of the target AI model to be issued;

and executing the AI model issuing instruction according to the first identification information and the second identification information so as to allow the target terminal equipment to receive and load the target AI model.

In a second aspect, an embodiment of the present invention further provides an AI model deployment method, which is applied to a display device, and the method includes:

displaying an AI model issuing page;

responding to the triggering operation of a user on the AI model issuing page, acquiring first identification information of target terminal equipment of the model to be deployed, and acquiring second identification information of the target AI model to be issued;

generating a model issuing instruction based on the first identification information and the second identification information;

and sending the model issuing instruction to a server for the server to execute the model issuing instruction so as to issue the target AI model corresponding to the second identification information to the target terminal equipment corresponding to the first identification information, so that the target terminal equipment receives and loads the target AI model.

In a third aspect, an embodiment of the present invention further provides a server, where the server includes a processor, a memory, and a computer program stored in the memory and executable by the processor, where the computer program, when executed by the processor, implements the steps of the AI model deployment method described above.

In a fourth aspect, an embodiment of the present invention further provides a display apparatus, where the display apparatus includes a display device, a processor, a memory, and a computer program stored on the memory and executable by the processor, where the computer program, when executed by the processor, implements the steps of the AI model deployment method as described above.

In a fifth aspect, an embodiment of the present invention further provides an AI model deployment system, where the AI model deployment system includes a terminal device, the server as described above, and the display device as described above, and the server is in communication connection with the display device and the terminal device, respectively.

In a sixth aspect, the present invention also provides a storage medium for a computer-readable storage, where the storage medium stores one or more programs, and the one or more programs are executable by one or more processors to implement the steps of the method for deploying an AI model according to any one of the embodiments provided in the present specification.

Based on the AI model deployment method, the server can respond to the user requirements, the initial AI model is trained according to the training data set provided by the user to obtain the target AI model with higher accuracy and matched with the user requirements, the user can issue a page through the AI model displayed by the display equipment, the model issuing instruction is triggered by one key and is sent to the server by the display equipment, and then the server issues the target AI model to the target terminal equipment based on the first identification information of the target terminal equipment of the model to be deployed and the second identification information of the target AI model to be issued in the model issuing instruction, so that the target terminal equipment can receive and load the target AI model. After the target AI model is obtained by automatically training the initial AI model, the target AI model is quickly and conveniently sent to the target terminal equipment through the AI model issuing page, so that the deployment flexibility and convenience of the AI model are greatly improved, and a user can quickly and conveniently use the AI model on the terminal.

Drawings

In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings required to be used in the description of the embodiments are briefly introduced below, and it is obvious that the drawings in the description below are some embodiments of the present application, and it is obvious for those skilled in the art to obtain other drawings without creative efforts.

Fig. 1 is a schematic view of a scenario for implementing an AI model deployment method according to an embodiment of the present invention;

fig. 2 is a schematic flowchart illustrating steps of an AI model deployment method according to an embodiment of the present invention;

FIG. 3 is a schematic diagram of a model issue page in an embodiment of the invention;

FIG. 4 is another diagram illustrating a model-issued sub-page in an embodiment of the present invention;

FIG. 5 is another diagram illustrating a model-issued sub-page in an embodiment of the present invention;

FIG. 6 is another diagram illustrating a model-issued sub-page in an embodiment of the present invention;

FIG. 7 is another schematic diagram of a model-issued sub-page in an embodiment of the present invention;

fig. 8 is a schematic flowchart illustrating steps of another AI model deployment method according to an embodiment of the present invention;

fig. 9 is a schematic flowchart illustrating steps of another AI model deployment method according to an embodiment of the present invention;

FIG. 10 is a block diagram illustrating a server according to an embodiment of the present invention;

fig. 11 is a block diagram schematically illustrating a structure of a display device according to an embodiment of the present invention;

fig. 12 is a block diagram illustrating a structure of an AI model deployment system according to an embodiment of the present invention.

Detailed Description

The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are some, but not all, embodiments of the present invention. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.

The flow diagrams depicted in the figures are merely illustrative and do not necessarily include all of the elements and operations/steps, nor do they necessarily have to be performed in the order depicted. For example, some operations/steps may be decomposed, combined or partially combined, so that the actual execution sequence may be changed according to the actual situation.

It is to be understood that the terminology used in the description of the invention herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention. As used in the specification of the present invention and the appended claims, the singular forms "a," "an," and "the" are intended to include the plural forms as well, unless the context clearly indicates otherwise.

Some embodiments of the invention are described in detail below with reference to the accompanying drawings. The embodiments described below and the features of the embodiments can be combined with each other without conflict.

With the rapid development of artificial intelligence, AI models are widely used in the fields of face recognition, image classification, object detection, speech recognition, natural language processing, and the like. At present, people mainly burn and write the AI model into terminal equipment through a data line or other modes, so that the terminal equipment can operate the AI model to perform face recognition, image classification, object detection, voice recognition and the like, but the AI model in the terminal equipment cannot be flexibly updated and replaced, and the user experience is poor.

In order to solve the above problems, embodiments of the present invention provide an AI model deployment method, a server, a display device, a system, and a storage medium, where based on the AI model deployment method, the server is capable of responding to a user demand, training an initial AI model according to a training data set provided by a user to obtain a target AI model with high accuracy and matching with the user demand, and then the user may issue a page through the AI model displayed by the display device, trigger a model issuing instruction with one key, and send the model issuing instruction to the server by the display device, and then issue the target AI model to the target terminal device based on first identification information of a target terminal device of the model to be deployed and second identification information of the target AI model to be issued in the model issuing instruction by the server, so that the target terminal device can receive and load the target AI model, and quickly and conveniently issue the target AI model to an underground target terminal device through the AI model issuing page, thereby greatly improving flexibility and convenience of the AI model deployment, and enabling the user to quickly and conveniently use the AI model on the terminal.

Referring to fig. 1, fig. 1 is a schematic view of a scenario for implementing an AI model deployment method according to an embodiment of the present invention. As shown in fig. 1, the scenario includes a server 100, a display device 200, and a terminal device 300, and the server 100 is communicatively connected to the display device 200 and the terminal device 300, respectively. The number of the terminal devices 300 connected to the server 100 may be multiple, and the server 100 may store multiple AI models, which may include a face recognition model, an image classification model, an object detection model, a high altitude parabolic detection model, a voice recognition model, a natural language processing model, and the like. The server 100 may also be configured to respond to a user demand, and perform model training on the initial AI model according to a training data set provided by the user, so as to obtain an AI model with higher accuracy and matching with the user demand, where the server may be a single server or a server cluster composed of multiple servers.

In an embodiment, the display device 200 includes a display device, and the display device is configured to display a model training page and a model issuing page of the server 100, where a user may perform model training work through operations on the model training page to generate a target AI model required by the user, and may deploy the target AI model to a corresponding terminal through operations on the model issuing page, where the terminal includes, but is not limited to, a mobile terminal, an edge device, and the like. It should be noted that the display device includes a display screen disposed on the display apparatus 200, and the display screen includes an LED display screen, an OLED display screen, an LCD display screen, and the like. The display device may be a smart phone, a tablet computer, a personal computer, a notebook computer, or the like, or may also be other electronic devices with a display screen, which is not specifically limited in this embodiment of the present invention.

In an embodiment, the terminal device 300 includes an image capturing device, through which a connection establishment page of the server 100 can be scanned to obtain an IP address of the server 100, so that the terminal device 300 can send a connection establishment request to the server 100 based on the IP address, and the server 100 establishes a communication connection between the terminal device 300 and the server 100 based on the connection establishment request sent by the terminal device 300. The connection establishment page includes a connection barcode image of the server, the connection barcode image carries an IP address of the server 100, the connection barcode image includes a one-dimensional code image or a two-dimensional code image, and the terminal device 300 may be a smart phone, a tablet computer, a personal computer, a notebook computer, a digital camera, a single lens reflex camera, a vehicle data recorder, a monitoring camera, a wearable device, a home appliance device, or the like.

In an embodiment, the display device 200 displays an AI model delivery page; the display device 200 responds to a trigger operation of a user on an AI model issuing page, acquires first identification information of a target terminal device of a model to be deployed, and acquires second identification information of the target AI model to be issued; generating a model issuing instruction based on the first identification information and the second identification information; the display device 200 sends a model issuing instruction to the server 100; the server 100 acquires a model issuing instruction sent by the display device 100, and acquires first identification information of a target terminal device of a model to be deployed and second identification information of a target AI model to be issued from the model issuing instruction; the server 100 executes an AI model issuing instruction based on the first identification information and the second identification information, so that the target terminal device receives and loads the target AI model.

Hereinafter, the AI model deployment method provided by the embodiment of the present invention will be described in detail with reference to the scenario in fig. 1. It should be noted that the scenario in fig. 1 is only used to explain the AI model deployment method provided by the embodiment of the present invention, but does not constitute a limitation on an application scenario of the AI model deployment method provided by the embodiment of the present invention.

Referring to fig. 2, fig. 2 is a schematic flowchart illustrating steps of an AI model deployment method according to an embodiment of the present invention. The AI model deployment method is applied to a server.

As shown in fig. 2, the AI model deployment method may include steps S101 to S105.

Step S101, a training data set and an initial AI model to be trained are obtained.

The server may store a plurality of initial AI models, which may include a face recognition model, an image classification model, an object detection model, a speech recognition model, a natural language processing model, and the like.

In one embodiment, a display device displays an AI model training page, and acquires a storage path of a training data set input by a user in the AI model training page and model identification information of an initial AI model to be trained; uploading the model identification information and the training data set under the storage path to a server; the server acquires the training data set and the model identification information uploaded by the display device, and determines the AI model corresponding to the model identification information as an initial AI model to be trained.

Illustratively, the AI model training page includes a data upload button, an AI model list, and a confirm button; acquiring model identification information of an initial AI model selected by a user in the AI model list; responding to the triggering operation of the data uploading key by the user, displaying a data uploading popup window, and acquiring a storage path input by the user in the data uploading popup window; and responding to the triggering operation of the user on the confirmation key, and uploading the model identification information and the training data set under the storage path to a server.

The initial AI model to be trained may also be obtained from the server by the server performing task judgment analysis on the uploaded training data set. That is, the user does not need to input the initial AI model to be trained, or select model identification information of the initial AI model to be trained, and the AI model corresponding to the model identification information may be determined as the initial AI model to be trained.

Step S102, training the initial AI model based on the training data set to generate a target AI model.

Illustratively, an automatic AI training platform is constructed in the server, and the automatic AI training platform can automatically train the initial AI model according to a training data set to obtain a target AI model with an ideal effect.

And step S103, receiving a model issuing instruction sent by the display equipment.

The model issuing instruction is generated by the display equipment according to the operation of the user on the displayed AI model issuing page. Illustratively, the model issuing instruction is generated by the display device according to the triggering operation of the user on the model issuing key in the displayed AI model issuing page. The model issues keys as virtual keys.

In one embodiment, the display device displays an AI model issuing page; responding to the triggering operation of a user on the AI model issuing page, acquiring first identification information of target terminal equipment of the model to be deployed, and acquiring second identification information of the target AI model to be issued; and generating a model issuing instruction based on the first identification information and the second identification information, and sending the model issuing instruction to the server. The AI model issuing page comprises a model issuing key, and the triggering operation of the user on the AI model issuing page comprises the triggering operation of the user on the model issuing key.

In one embodiment, the model issue page includes an AI model information bar and a model issue button. It can be understood that one model issuing key may correspond to one AI model information field, that is, one model issuing key issues one AI model, and one model issuing key may also correspond to a plurality of AI model information fields, that is, one model issuing key issues a plurality of AI models. The AI model information column may include, but is not limited to, a model name, a training task progress, an update time, a data source, an accuracy rate, a training detail button, a model prediction button, a model issue button, and the like of the AI model.

Exemplarily, in response to a triggering operation of a user on the AI model issuing key, each terminal device connected with the server is determined as a target terminal device of the model to be deployed, and the AI model corresponding to the AI model issuing key is determined as a target AI model to be issued; and acquiring first identification information of each target terminal device and second identification information of each target AI model.

For example, as shown in fig. 3, the model delivery page includes 4 AI model information fields, which are an AI model information field 11, an AI model information field 12, an AI model information field 13, and an AI model information field 14, respectively, where the model name, the training task progress, the update time, the data source, and the accuracy of the AI model corresponding to the AI model information field 11 are "2021-05-07-auto training-headgear wearing data set", "training success", "2021-5-7" 23 52"," headgear wearing data set-v 1", and 91%, respectively. The model name, training task progress, update time, data source, and accuracy of the AI model corresponding to the AI model information column 12 are "2021-05-07-auto training-headgear wearing data set", "training success", "2021-5-10.

The model name, the training task progress, the update time, the data source, and the accuracy of the AI model corresponding to the AI model information column 13 are "2021-04-30-automatic training-headgear wearing data set", "training success", "2021-4-30" 27"," headgear wearing data set-v 1", and 89%, respectively. The model name, the training task progress, the update time, the data source, and the accuracy of the AI model corresponding to the AI model information column 14 are "2021-04-30-automatic training-fan-v 1-model", "training success", "2021-4-30 16", 48"," fan-v 1", and 98%, respectively. If the triggering operation of the user on the model issuing key in the AI model information column 11 is detected, the AI model corresponding to the AI model information column 11 is determined as the AI model and determined as the target AI model to be issued.

In one embodiment, the display device displays an AI model issuing page, wherein the AI model issuing page comprises an AI model information bar and a model issuing key; responding to the triggering operation of the user on the model issuing key, and displaying an AI model issuing sub-page corresponding to the triggered model issuing key; responding to the triggering operation of a user on an AI model issuing sub-page, acquiring first identification information of target terminal equipment of a model to be deployed, and acquiring second identification information of the target AI model to be issued; and generating a model issuing instruction based on the first identification information and the second identification information, and sending the model issuing instruction to the server.

In one embodiment, the AI model issuing sub-page comprises a model issuing key, at least one terminal information bar and a selection key corresponding to the terminal information bar, responds to the triggering operation of a user on the model issuing key, and determines first identification information of target terminal equipment of a model to be deployed according to the state information of each selection key; and determining the AI model corresponding to the AI model issuing key as a target AI model to be issued, and acquiring second identification information of each target AI model. The state information corresponding to the selection key comprises first state information or second state information, the first state information is used for indicating that the selection key is in an open state, and the second state information is used for indicating that the selection key is in a closed state.

The terminal information bar may include, but is not limited to, a device ID, a device name, a registration code validity period, and device status information of the terminal device, where the device status information includes offline status information or online status information, the online status information is used to indicate that an application currently running by the terminal device is a preset application, the terminal device is in an online status, the offline status information is used to indicate that the application currently running by the terminal device is not the preset application, and the terminal device is in an offline status. The preset application program is used for realizing data interaction between the server and the terminal equipment.

For example, as shown in fig. 4, the model-issued sub-page is located in the lower right corner of the model-issued sub-page, and the device IDs of these 5 Terminal devices are respectively 13.

The registration code validity period is a validity period of the device ID determined when the terminal device registers the account in the server, and is used for managing the validity period of the device ID of the terminal device, that is, after the device ID of the terminal device exceeds the validity period, the terminal device needs to register a new device ID with the server again, and determine the validity period of the new registration code. The validity period of the registration code may be set based on actual conditions, which is not specifically limited in the embodiment of the present invention.

In one embodiment, the display mode of the triggered selection key is switched and the state information corresponding to the triggered selection key is updated in response to the triggering operation of a user on the selection key; responding to the triggering operation of a user on the AI model issued keys, and determining first identification information of the target terminal equipment of the model to be deployed according to the updated state information corresponding to each selection key; and determining the AI model corresponding to the AI model issuing key as a target AI model to be issued, and acquiring second identification information of each target AI model. The user can determine the target terminal equipment of the model to be deployed through the triggering operation of the selection or adjustment key, so that the convenience of deploying the AI model is improved.

The display mode of the selection key comprises a first display mode or a second display mode, the first display mode is different from the second display mode, the first display mode is used for showing that the selection key is in an open state, namely, a user selects the corresponding terminal device as the target terminal device of the model to be deployed, and the second display mode is used for showing that the selection key is in a closed state, namely, the user does not select the corresponding terminal device as the target terminal device of the model to be deployed. For example, the first display mode is that the display color of the selection key is black, and the second display mode is that the display color of the selection key is white.

When detecting a triggering operation of the user on the selection key corresponding to device ID 57.

In an embodiment, a user may issue an AI model to be deployed of a terminal device corresponding to a sub-page configuration terminal information field through a model, and may configure one or more AI models to be deployed for the terminal device, specifically: the display equipment responds to the triggering operation of a user on the terminal information bar and displays an AI model pull-down menu of the corresponding terminal equipment, wherein the AI model pull-down menu comprises an AI model deleting key and an AI model adding key; responding to the triggering operation of a user on an AI model adding key, and displaying an AI model selection popup window, wherein the AI model selection popup window comprises model IDs of a plurality of AI models and a confirmation key; acquiring a target model ID selected by a user in the AI model selection popup, and when the triggering operation of the user on the confirmation key is detected, newly adding a corresponding AI model option bar in the AI model pull-down menu according to the target model ID; and responding to the triggering operation of the user on the terminal equipment option displaying the AI model pull-down menu, hiding the AI model pull-down menu, and generating AI model configuration information corresponding to the terminal equipment.

In an embodiment, the manner of obtaining the second identification information of the AI model to be delivered may be: determining first identification information of target terminal equipment of the model to be deployed according to the updated state information corresponding to each selection key; and obtaining AI model configuration information of the target terminal equipment corresponding to the first identification information, and obtaining second identification information of the target AI model to be issued from the AI model configuration information.

For example, as shown in fig. 4, when the user clicks the second terminal information field with a finger or a mouse, the displayed AI model drop-down menu is as shown in fig. 6, the AI model drop-down menu displays an AI model option field, and the AI model option field displays model ID, model name, and accuracy of 2m. As shown in fig. 7, the model ID, the model name, and the accuracy rate displayed in the newly added AI model option column are 52 c 4 c.

And step S104, analyzing the model issuing command to acquire first identification information of the target terminal equipment of the model to be deployed and second identification information of the target AI model to be issued.

The model issuing instruction may include first identification information of one target terminal device and second identification information of one target AI model, or may also include first identification information of a plurality of target terminal devices and second identification information of a plurality of target AI models, or may also include first identification information of one target terminal device and second identification information of a plurality of target AI models, which is not specifically limited in this embodiment of the present invention.

For example, the first identification information is used to uniquely identify the target terminal device, the first identification information may include a device ID and/or a device name of the target terminal device, the second identification information is used to uniquely identify the target AI model, the second identification information may include a model ID and/or a model name of the target AI model, and characters in the first identification information or the second identification information may include numbers, large and small letters, lower case letters, and/or greek letters, and may also include the remaining characters, which is not specifically limited in this embodiment of the present invention.

And S105, executing an AI model issuing instruction according to the first identification information and the second identification information so as to allow the target terminal equipment to receive and load the target AI model.

For example, if there is one first identifier and one second identifier, the target AI model corresponding to the second identifier is sent to the target terminal device corresponding to the first identifier.

And if the first identification information is multiple and the second identification information is one, issuing the target AI model corresponding to the second identification information to the target terminal equipment corresponding to each first identification information, namely issuing one target AI model to multiple target terminal equipment.

If the first identification information is one and the second identification information is multiple, all the target AI models corresponding to each second identification information are sent to the target terminal equipment corresponding to the first identification information, that is, all the target AI models are sent to one target terminal equipment.

In an embodiment, in the process of issuing the target AI model to the target terminal device, the server sends the downloading progress information of the target AI model to the target terminal device, and the target terminal device displays the downloading progress information of the target AI model. Wherein, the download progress information comprises a download percentage and a download progress bar. Through the download progress information of the target AI model, the user can know the download progress of the target AI model conveniently, and the user experience is improved.

In an embodiment, the model issuing instruction carries a plurality of first identification information, a plurality of second identification information, and mapping relationship information between the first identification information and the second identification information, and therefore, executing the AI model issuing instruction according to the first identification information and the second identification information may include: determining at least one piece of second identification information corresponding to each piece of first identification information from a plurality of pieces of second identification information according to the mapping relation information; and issuing at least one corresponding target AI model to the target terminal equipment corresponding to each first identification information according to at least one second identification information corresponding to each first identification information. Through the scheme, the same or different AI models can be deployed on different terminal devices, and one terminal device can deploy one or more AI models, so that the terminal device can run one or more AI models.

The model issuing instruction carries a plurality of first identification information, a plurality of second identification information, and mapping relation information between the first identification information and the second identification information, wherein the at least one second identification information corresponding to each first identification information can be the same or different, and the mapping relation information is configured by a user through the model issuing page. For example, there are 3 pieces of first identification information, namely 13.

TABLE 1

First identification information	Second identification information
		13:26:14:52:94:85	A1:48:B5:49、57:R5:5T:20
46:68:39:62:85:73	52:4C:V5:96、2M:58:D9:93
		57:75:01:29:55:20	2M:58:D9:93、52:4C:V5:96、A1:48:B5:49

As shown in table 1, the target AI model 52.

In an embodiment, because the computing power of the terminal device is limited and some AI models have a high demand on the computing power, under the condition that the computing power of the terminal device is low, the terminal device cannot effectively deploy and run the AI models having a high demand on the computing power, and therefore, the computing power index of the target terminal device corresponding to the first identification information is determined; if the determined calculation capacity index is smaller than or equal to a preset calculation capacity index threshold, compressing a target AI model corresponding to the second identification information; and issuing the compressed target AI model to the target terminal equipment corresponding to the first identification information so that the target terminal equipment can receive and load the compressed target AI model. And after the target AI model is compressed, the compressed target AI model is issued to the terminal equipment, so that the terminal equipment can conveniently operate and deploy the AI model.

For example, the compressing the target AI model corresponding to the second identification information may be: and determining a compression strategy of the target AI model according to the calculation capacity index of the target terminal equipment, and compressing the target AI model according to the determined compression strategy. The server stores a mapping relation between the computing power index and the compression strategy, and the compression strategy of the target AI model can be determined in a self-adaptive manner through the mapping relation and the computing power index of the target terminal equipment, so that the target AI model compressed according to the determined compression strategy can be better adapted to the target terminal equipment, the flexibility of AI model deployment can be further improved, and the terminal equipment can conveniently operate and deploy the AI model.

For example, the manner of determining the computing capability index of the target terminal device corresponding to the first identification information may be: and acquiring a mapping relation between pre-stored identification information and the calculation capacity index, and determining the calculation capacity index of the target terminal device corresponding to the first identification information according to the mapping relation and the first identification information. The computing power index of the terminal device is used for representing the computing power of the terminal device, the higher the computing power index is, the stronger the computing power of the terminal device is, and the lower the technical power index is, the weaker the computing power of the terminal device is. The mapping relationship between the pre-stored identification information and the calculation capability index is determined according to the chip type and the identification information of the terminal device establishing communication connection with the server, and the preset calculation capability index threshold value may be set based on an actual situation, which is not specifically limited in the embodiment of the present application.

In an embodiment, determining type information of a target terminal device corresponding to first identification information; determining a target format of the target AI model according to the type information, and converting the format of the target AI model corresponding to the second identification information into the target format; and issuing the target AI model after format conversion to the target terminal equipment corresponding to the first identification information so as to be received by the target terminal equipment and loaded with the compressed target AI model. The server stores a mapping relation between the type information and the format, and can determine the target format of the target AI model according to the mapping relation and the type information of the target terminal equipment. The AI model can be conveniently operated and deployed by the terminal equipment through self-adaptive conversion of the format of the AI model based on the type information of the terminal equipment.

In an embodiment, a model loading list of the target terminal device corresponding to the first identification information is obtained, where the model loading list includes third identification information of an AI model that has been loaded by the target terminal device; and if the second identification information is different from the third identification information, issuing the target AI model corresponding to the second identification information to the target terminal equipment corresponding to the first identification information, so that the target terminal equipment receives and loads the target AI model to update the loaded AI model. By the mode, the AI model deployed in the terminal equipment can be updated based on the requirement of the user, so that the deployed AI model is more suitable for the requirement of the user.

In an embodiment, if the second identification information is the same as the third identification information, determining whether the second version information of the target AI model corresponding to the second identification information is different from the first version information of the loaded AI model in the model loading list; and if the second version information is different from the first version information, issuing the target AI model corresponding to the second identification information to the target terminal equipment corresponding to the first identification information, so that the target terminal equipment can receive and load the target AI model to update the loaded AI model. By the aid of the scheme, the AI model deployed in the terminal equipment can be updated timely after the AI model in the server is updated.

In an embodiment, if the second version information is the same as the first version information, the target AI model corresponding to the second identification information is not issued to the target terminal device corresponding to the first identification information. The same AI model is not issued to the terminal equipment under the condition that the issued AI model is the same as the loaded AI model in the terminal equipment, so that the repeated deployment of the same AI model in the terminal equipment is avoided. In another embodiment, when receiving a target AI model delivered by a server, a terminal device sends delivery interruption information to the server if it is determined that the delivered target AI model is the same as a locally loaded AI model, and when receiving the delivery interruption information, the server stops delivering the target AI model to the terminal device. Through the scheme, the repeated deployment of the same AI model in the terminal equipment can be avoided.

In one embodiment, a connection establishment page between a terminal device and a server is displayed; the terminal equipment scans a connection establishment page to obtain an IP address of the server, generates a connection establishment request according to the IP address and then sends the connection establishment request to the server; the server acquires a connection establishment request sent by the terminal equipment, and establishes communication connection between the terminal equipment and the server according to the connection establishment request. The page is established by the connection of the terminal equipment scanning server, so that the communication connection between the server and the terminal equipment can be conveniently established, and the user experience is improved.

The connection establishment request is generated by scanning a connection establishment page of the server by the terminal equipment, the connection establishment page comprises a connection barcode image of the server, the connection barcode image carries an IP address of the server, and the connection barcode image comprises a one-dimensional code image or a two-dimensional code image. Illustratively, the terminal device scans a connection barcode image in the connection establishment page through the image acquisition device to obtain an IP address of the server, and sends a connection establishment request to the server according to the IP address. The terminal equipment scans the one-dimensional code image or the two-dimensional code image of the server, so that the communication connection between the server and the terminal equipment can be conveniently established, and the user experience is improved.

In one embodiment, after the terminal device establishes communication connection with the server, the server acquires attribute information of the terminal device, generates a list updating instruction according to the attribute information, and then sends the list updating instruction to the display device; the display equipment acquires a list updating instruction sent by the server, acquires attribute information of new terminal equipment which establishes connection with the server from the list updating instruction, and updates the terminal equipment connection list according to the attribute information.

The attribute information comprises an equipment ID, an equipment name, a registration code validity period, equipment state information and a terminal upper model list of the terminal equipment, after the communication connection between the terminal equipment and the server is established, the equipment ID and the validity period are distributed to the terminal equipment, the terminal equipment runs a preset application program, and in the running preset application program, an IP address of the server obtained by scanning the target two-dimensional code by the image acquisition device is obtained, so that a connection establishment request is sent to the server according to the IP address.

In an embodiment, if a plurality of target AI models are available, a data reception capability index of the target terminal device corresponding to the first identification information is obtained; determining the issuing sequence of each target AI model according to the data capacity receiving index and the data quantity of each target AI model; and issuing each target AI model to the target terminal equipment corresponding to the first identification information according to the issuing sequence of each target AI model. The data receiving capability index of the target terminal device is used for representing the maximum data volume of the data which can be received by the target terminal device, the higher the receiving capability index is, the more the maximum data volume of the data which can be received by the target terminal device is, and the lower the receiving capability index is, the less the maximum data volume of the data which can be received by the target terminal device is. By the scheme, data transmission congestion between the server and the target terminal equipment can be reduced, data transmission efficiency is greatly improved, and terminal equipment can conveniently deploy the AI model.

For example, according to the data capacity receiving index and the data amount of each target AI model, the manner of determining the issuing order of each target AI model may be: acquiring data transmission quantity corresponding to the data capacity receiving index; determining a ratio between the data amount of each target AI model and the data transmission amount; and determining the issuing sequence of each target AI model according to the ratio of the data volume of each target AI model to the data transmission volume. For example, the ratios between the data amounts of the target AI model a, the target AI model B, and the target AI model C and the data transmission amount are 0.4, 0.5, and 0.9, respectively, the following sequence of the target AI model a, the target AI model B, and the target AI model C is: the target AI model A is issued first, then the target AI model B is issued, and finally the target AI model C is issued, or the target AI model A and the target AI model are issued first and then the target AI model C is issued, or the target AI model C is issued first and then the target AI model A and the target AI model are issued simultaneously.

In an embodiment, the model issuing instruction further carries a calling sequence of the plurality of target AI models, so that after the target terminal device receives and loads the plurality of target AI models, the target AI models can sequentially execute the multi-model task based on the calling sequence of the plurality of target AI models. For example, the target AI model D, the target AI model E, and the target AI model F are called in the order of first calling the target AI model E, then calling the target AI model F, and finally calling the target AI model D, so that after the target terminal device receives and loads the target AI model D, the target AI model E, and the target AI model F, first calling the target AI model E, then calling the target AI model F, and finally calling the target AI model D to complete the multi-model task.

In the AI model deployment method provided in the above embodiment, the server can respond to the user requirement, train the initial AI model according to the training data set provided by the user to obtain the target AI model with higher accuracy and matching with the user requirement, then the user can issue a page through the AI model displayed by the display device, trigger the model issuing instruction by one key, send the model issuing instruction to the server by the display device, and then issue the target AI model to the target terminal device by the server based on the first identification information of the target terminal device of the model to be deployed and the second identification information of the target AI model to be issued in the model issuing instruction, so that the target terminal device can receive and load the target AI model, thereby greatly improving the deployment flexibility and convenience of the AI model in the terminal device, and improving the user experience.

Referring to fig. 8, fig. 8 is a schematic flowchart illustrating steps of another AI model deployment method according to an embodiment of the present invention. The AI model deployment method is applied to display equipment.

As shown in fig. 8, the AI model deployment method may include steps S201 to S204.

Step S201, displaying an AI model issuing page;

step S202, responding to the triggering operation of a user on an AI model issuing page, acquiring first identification information of a target terminal device of a model to be deployed, and acquiring second identification information of the target AI model to be issued;

the AI model issuing page comprises a model issuing key, and the triggering operation of the user on the AI model issuing page comprises the triggering operation of the user on the model issuing key. Exemplarily, in response to a trigger operation of a user on the AI model issuing key, determining each terminal device connected to the server as a target terminal device of the model to be deployed, and determining the AI model corresponding to the AI model issuing key as a target AI model to be issued; and acquiring first identification information of each target terminal device and second identification information of each target AI model.

In one embodiment, the model issue page includes an AI model information bar and a model issue button. It can be understood that one model issuing key may correspond to one AI model information field, that is, one model issuing key issues one AI model, and one model issuing key may also correspond to a plurality of AI model information fields, that is, one model issuing key issues a plurality of AI models. The AI model information column comprises the model name of the AI model, the progress of a training task, the updating time, the data source, the accuracy, a training detail key, a model prediction key, a model issuing key and the like.

In one embodiment, the display device displays an AI model issuing page, wherein the AI model issuing page comprises an AI model information bar and a model issuing key; responding to the triggering operation of the user on the model issuing key, and displaying an AI model issuing sub-page corresponding to the triggered model issuing key; responding to the triggering operation of a user on an AI model issuing sub-page, acquiring first identification information of target terminal equipment of a model to be deployed, and acquiring second identification information of the target AI model to be issued; and generating a model issuing instruction based on the first identification information and the second identification information, and sending the model issuing instruction to a server.

In one embodiment, the AI model issuing sub-page comprises a model issuing key, at least one terminal information bar and a selection key corresponding to the terminal information bar, responds to the triggering operation of a user on the model issuing key, and determines first identification information of target terminal equipment of a model to be deployed according to the state information of each selection key; and determining the AI model corresponding to the AI model issuing key as a target AI model to be issued, and acquiring second identification information of each target AI model. The state information corresponding to the selection key comprises first state information or second state information, wherein the first state information is used for indicating that the selection key is in an open state, and the second state information is used for indicating that the selection key is in a closed state.

Step S203, generating a model issuing command based on the first identification information and the second identification information;

and step S204, sending a model issuing instruction to the server so that the server can execute the model issuing instruction to issue the target AI model corresponding to the second identification information to the target terminal equipment corresponding to the first identification information, so that the target terminal equipment can receive and load the target AI model.

In one embodiment, a connection establishment page between a terminal device and a server is displayed, and the terminal device generates a connection establishment request and sends the connection establishment request to the server after scanning the connection establishment page; acquiring a list updating instruction sent by a server, wherein the list updating instruction is generated after the server establishes connection with new terminal equipment based on a connection establishment request sent by the terminal equipment; and acquiring the attribute information of the new terminal equipment from the list updating instruction, and updating the terminal equipment connection list according to the attribute information. The connection establishment page comprises a connection barcode image of the server, the connection barcode image carries an IP address of the server, the connection barcode image comprises a one-dimensional code image or a two-dimensional code image, the terminal equipment connection list comprises attribute information of terminal equipment connected with the server, and the attribute information comprises an equipment ID, an equipment name, a registration code validity period, equipment state information and a terminal model list of the terminal equipment. The page is established by the connection of the terminal equipment scanning server, so that the communication connection between the server and the terminal equipment can be conveniently established, and the user experience is improved.

In the AI model deployment method provided in the above embodiment, a user may send a model issuing instruction to a server by one key through an AI model issuing page displayed by a display device, and the server issues a target AI model to the target terminal device based on first identification information of a target terminal device of the model to be deployed and second identification information of the target AI model to be issued in the model issuing instruction, so that the target terminal device can receive and load the target AI model, thereby greatly improving deployment flexibility and convenience of the AI model in the terminal device, and improving user experience.

Referring to fig. 9, fig. 9 is a schematic flowchart illustrating steps of another AI model deployment method according to an embodiment of the present invention.

As shown in fig. 9, the AI model deployment method may include steps S301 to S305.

And S301, displaying an AI model issuing page by the display equipment.

Illustratively, the model issue page includes an AI model information bar and a model issue button. It can be understood that one model issuing key may correspond to one AI model information field, that is, one model issuing key issues one AI model, and one model issuing key may also correspond to a plurality of AI model information fields, that is, one model issuing key issues a plurality of AI models. The AI model information column may include, but is not limited to, a model name, a training task progress, an update time, a data source, an accuracy rate, a training detail button, a model prediction button, a model issue button, and the like of the AI model.

Step S302, the display device responds to the trigger operation of the user to the AI model issuing page, obtains the first identification information of the target terminal device of the model to be deployed, and obtains the second identification information of the target AI model to be issued.

And step S303, the display device generates a model issuing instruction based on the first identification information and the second identification information, and sends the model issuing instruction to the server.

Step S304, the server receives the model issuing command sent by the display device and analyzes the model issuing command to obtain the first identification information of the target terminal device of the model to be deployed and the second identification information of the target AI model to be issued.

For example, the first identification information is used to uniquely identify the target terminal device, the first identification information may include a device ID and/or a device name of the target terminal device, the second identification information is used to uniquely identify the target AI model, the second identification information may include a model ID and/or a model name of the target AI model, and characters in the first identification information or the second identification information may include numbers, large and small letters, lower case letters, and/or greek letters, and of course, other characters may also be included, which is not specifically limited in this embodiment of the present invention.

Step S305, the server executes an AI model issuing instruction according to the first identification information and the second identification information, so that the target terminal equipment can receive and load the target AI model.

Illustratively, if the first identification information is one and the second identification information is one, the target AI model corresponding to the second identification information is issued to the target terminal device corresponding to the first identification information. And if the first identification information is multiple and the second identification information is one, issuing the target AI model corresponding to the second identification information to the target terminal equipment corresponding to each first identification information, namely issuing one target AI model to multiple target terminal equipment. If the first identification information is one and the second identification information is multiple, all the target AI models corresponding to each second identification information are sent to the target terminal equipment corresponding to the first identification information, that is, all the target AI models are sent to one target terminal equipment. If the first identification information is multiple and the second identification information is multiple, all the target AI models corresponding to each second identification information are issued to the target terminal equipment corresponding to each first identification information, that is, all the target AI models are issued to each target terminal equipment.

It should be noted that, as will be clear to those skilled in the art, for convenience and brevity of description, the specific working process of the present embodiment may refer to the corresponding process in the foregoing embodiment, and is not described herein again.

Referring to fig. 10, fig. 10 is a schematic block diagram of a server according to an embodiment of the present invention.

As shown in fig. 10, the server 400 includes a processor 410 and a memory 420, and the processor 410 and the memory 420 are connected by a bus 430, such as an I2C (Inter-integrated Circuit) bus.

In particular, processor 410 is configured to provide computational and control capabilities to support the operation of the overall server. The Processor 410 may be a Central Processing Unit (CPU), and the Processor 410 may also be other general purpose processors, digital Signal Processors (DSPs), application Specific Integrated Circuits (ASICs), field Programmable Gate Arrays (FPGAs) or other Programmable logic devices, discrete Gate or transistor logic devices, discrete hardware components, etc. Wherein a general purpose processor may be a microprocessor or the processor may be any conventional processor or the like.

Specifically, the Memory 420 may be a Flash chip, a Read-Only Memory (ROM) magnetic disk, an optical disk, a usb disk, or a removable hard disk.

Those skilled in the art will appreciate that the architecture shown in fig. 10 is merely a block diagram of some of the structures associated with an embodiment of the present invention and does not constitute a limitation on the servers to which an embodiment of the present invention may be applied, and that a particular server may include more or less components than those shown, or some of the components may be combined, or have a different arrangement of components.

In an embodiment, the processor is configured to run a computer program stored in the memory and to implement the following steps when executing the computer program:

acquiring a training data set and an initial AI model to be trained;

In an embodiment, when implementing that the AI model issues an instruction according to the first identification information and the second identification information, the processor is configured to implement:

obtaining a model loading list of the target terminal device corresponding to the first identification information, wherein the model loading list comprises third identification information of an AI model loaded by the target terminal device;

and if the second identification information is different from the third identification information, issuing the target AI model corresponding to the second identification information to the target terminal equipment corresponding to the first identification information, so that the target terminal equipment receives and loads the target AI model to update the loaded AI model of the target terminal equipment.

In an embodiment, the model load list further includes first version information of loaded AI models, and the processor is further configured to:

if the second identification information is the same as the third identification information, determining whether second version information of the target AI model corresponding to the second identification information is different from the first version information;

if the second version information is different from the first version information, issuing a target AI model corresponding to the second identification information to a target terminal device corresponding to the first identification information, so that the target terminal device receives and loads the target AI model to update the loaded AI model of the target terminal device;

and if the second version information is the same as the first version information, not issuing a target AI model corresponding to the second identification information to the target terminal equipment corresponding to the first identification information.

In one embodiment, the processor is further configured to implement the steps of:

acquiring a connection establishment request sent by terminal equipment, wherein the connection establishment request is generated by scanning a connection establishment page of the server by the terminal equipment;

and establishing communication connection between the terminal equipment and the server according to the connection establishment request.

determining a computing capacity index of the target terminal device corresponding to the first identification information;

if the computing power index is smaller than or equal to a preset computing power index threshold value, compressing the target AI model corresponding to the second identification information;

and issuing the compressed target AI model to the target terminal equipment corresponding to the first identification information so that the target terminal equipment can receive and load the compressed target AI model.

In an embodiment, the target AI model is a plurality of models, and the processor is further configured to:

acquiring a data receiving capacity index of the target terminal device corresponding to the first identification information;

determining the issuing sequence of each target AI model according to the data capacity receiving index and the data volume of each target AI model;

and issuing each target AI model to the target terminal equipment corresponding to the first identification information according to the issuing sequence of each target AI model.

It should be noted that, as is clear to those skilled in the art, for convenience and simplicity of description, the specific working process of the server described above may refer to the corresponding process in the foregoing embodiment of the AI model deployment method, and details are not described herein again.

Referring to fig. 11, fig. 11 is a schematic block diagram of a display device according to an embodiment of the present invention.

As shown in fig. 11, the display apparatus 500 includes a display device 510, a processor 520, and a memory 530, and the display device 510, the processor 520, and the memory 530 are connected by a bus 540, such as an I2C (Inter-integrated Circuit) bus 540.

In particular, processor 520 is configured to provide computational and control capabilities to support the operation of the overall server. The Processor 520 may be a Central Processing Unit (CPU), and the Processor 520 may be other general purpose processors, digital Signal Processors (DSPs), application Specific Integrated Circuits (ASICs), field Programmable Gate Arrays (FPGAs) or other Programmable logic devices, discrete Gate or transistor logic devices, discrete hardware components, etc. Wherein a general purpose processor may be a microprocessor or the processor may be any conventional processor or the like.

Specifically, the Memory 530 may be a Flash chip, a Read-Only Memory (ROM) magnetic disk, an optical disk, a usb disk, or a removable hard disk.

It will be understood by those skilled in the art that the structure shown in fig. 11 is a block diagram of only a portion of the structure associated with an embodiment of the present invention, and does not constitute a limitation on the display device to which an embodiment of the present invention is applied, and a particular display device may include more or less components than those shown in the drawings, or may combine some components, or have a different arrangement of components.

displaying an AI model issuing page through a display device;

In an embodiment, the AI model delivery page includes a model delivery key, and the processor is configured to implement, when implementing a trigger operation of a response user on the AI model delivery page, acquiring first identification information of a target terminal device of a model to be deployed and acquiring second identification information of a target AI model to be delivered:

responding to the triggering operation of a user on the model issuing key, and determining each terminal device connected with the server as a target terminal device of the model to be deployed;

determining an AI model corresponding to the triggered AI model issuing key as a target AI model to be issued;

and acquiring first identification information of each target terminal device and second identification information of each target AI model.

displaying a connection establishment page between a terminal device and a server through a display device, generating a connection establishment request after the terminal device scans the connection establishment page, and sending the connection establishment request to the server;

acquiring a list updating instruction sent by the server, wherein the list updating instruction is generated after the server establishes communication connection with new terminal equipment based on the connection establishment request;

and acquiring the attribute information of the new terminal equipment from the list updating instruction, and updating the terminal equipment connection list according to the attribute information.

It should be noted that, as will be clearly understood by those skilled in the art, for convenience and brevity of description, the specific working process of the display device described above may refer to the corresponding process in the foregoing embodiment of the AI model deployment method, and details are not described here again.

Referring to fig. 12, fig. 12 is a schematic block diagram illustrating a structure of an AI model deployment system according to an embodiment of the present invention.

As shown in fig. 12, the AI model deployment system 600 includes a server 610, a display device 620, and a terminal device 630, and the server 610 is communicatively connected to the display device 620 and the terminal device 630, respectively. The server 610 may be the server 400 shown in fig. 10, and the display device 620 may be the display device 500 shown in fig. 11.

It should be noted that, as will be clearly understood by those skilled in the art, for convenience and brevity of description, the specific working process of the AI model deployment system described above may refer to the corresponding process in the foregoing AI model deployment method embodiment, and details are not described herein again.

Embodiments of the present invention also provide a storage medium for a computer-readable storage, the storage medium storing one or more programs, the one or more programs being executable by one or more processors to implement the steps of any one of the methods for deploying an AI model as provided in the specification of the embodiments of the present invention.

The storage medium may be an internal storage unit of the server, the display device, or the terminal device described in the foregoing embodiment, for example, a hard disk or a memory of the server, the display device, or the terminal device. The storage medium may also be an external storage device of the server, the display device or the terminal device, for example, a plug-in hard disk, a Smart Media Card (SMC), a Secure Digital (SD) Card, a Flash memory Card (Flash Card) or the like provided on the server, the display device or the terminal device.

It will be understood by those of ordinary skill in the art that all or some of the steps of the methods, systems, functional modules/units in the devices disclosed above may be implemented as software, firmware, hardware, or suitable combinations thereof. In a hardware embodiment, the division between functional modules/units mentioned in the above description does not necessarily correspond to the division of physical components; for example, one physical component may have multiple functions, or one function or step may be performed by several physical components in cooperation. Some or all of the physical components may be implemented as software executed by a processor, such as a central processing unit, digital signal processor, or microprocessor, or as hardware, or as an integrated circuit, such as an application specific integrated circuit. Such software may be distributed on computer readable media, which may include computer storage media (or non-transitory media) and communication media (or transitory media). The term computer storage media includes volatile and nonvolatile, removable and non-removable media implemented in any method or technology for storage of information such as computer readable instructions, data structures, program modules or other data, as is well known to those skilled in the art. Computer storage media includes, but is not limited to, RAM, ROM, EEPROM, flash memory or other memory technology, CD-ROM, digital Versatile Disks (DVD) or other optical disk storage, magnetic cassettes, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other medium which can be used to store the desired information and which can accessed by a computer. In addition, communication media typically embodies computer readable instructions, data structures, program modules or other data in a modulated data signal such as a carrier wave or other transport mechanism and includes any information delivery media as is well known to those skilled in the art.

It should be understood that the term "and/or" as used in this specification and the appended claims refers to and includes any and all possible combinations of one or more of the associated listed items. It should be noted that, in this document, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or system that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or system. Without further limitation, an element defined by the phrases "comprising one of 8230; \8230;" 8230; "does not exclude the presence of additional like elements in a process, method, article, or system that comprises the element.

The above-mentioned serial numbers of the embodiments of the present invention are only for description, and do not represent the advantages and disadvantages of the embodiments. While the invention has been described with reference to specific embodiments, the invention is not limited thereto, and various equivalent modifications and substitutions can be easily made by those skilled in the art without departing from the scope of the invention. Therefore, the protection scope of the present invention shall be subject to the protection scope of the claims.

Claims

1. An AI model deployment method, applied to a server, the method comprising:

acquiring a training data set and an initial AI model to be trained;

2. The AI model deployment method of claim 1, wherein the executing the AI model delivery instruction according to the first identification information and the second identification information comprises:

3. The AI model deployment method of claim 2, wherein the model loading list further includes first version information of the loaded AI model, the method further comprising:

4. The AI model deployment method of any of claims 1-3, wherein the method further comprises:

and establishing the communication connection between the terminal equipment and the server according to the connection establishment request.

5. The AI-model deployment method of any of claims 1-3, wherein the executing the AI-model issuing instruction according to the first identification information and the second identification information comprises:

determining a computing capacity index of the target terminal equipment corresponding to the first identification information;

if the computing power index is smaller than or equal to a preset computing power index threshold, compressing a target AI model corresponding to the second identification information;

6. The AI model deployment method of any of claims 1-3, wherein the target AI model is a plurality of, the method further comprising:

7. An AI model deployment method is applied to a display device, and the method comprises the following steps:

displaying an AI model issuing page;

and sending the model issuing instruction to a server so that the server can execute the model issuing instruction, and issuing the target AI model corresponding to the second identification information to the target terminal equipment corresponding to the first identification information so that the target terminal equipment can receive and load the target AI model.

8. The AI model deployment method of claim 7, wherein the AI model delivery page includes a model delivery button, and the acquiring a first identification information of a target terminal device of a model to be deployed and a second identification information of a target AI model to be delivered in response to a trigger operation of a user on the AI model delivery page includes:

9. The AI model deployment method of claim 7, further comprising:

displaying a connection establishment page between a terminal device and a server, wherein the terminal device generates a connection establishment request after scanning the connection establishment page and sends the connection establishment request to the server;

acquiring a list updating instruction sent by the server, wherein the list updating instruction is generated after the server establishes communication connection with new terminal equipment based on the connection establishing request;

10. A server, characterized in that the server comprises a processor, a memory and a computer program stored on the memory and executable by the processor, wherein the computer program, when executed by the processor, implements the steps of the AI model deployment method of any of claims 1 to 6.

11. A display device, characterized in that the display device comprises a display means, a processor, a memory, and a computer program stored on the memory and executable by the processor, wherein the computer program, when executed by the processor, implements the steps of the AI model deployment method of any of claims 7 to 9.

12. AI-model deployment system characterized in that it comprises a terminal device, a server according to claim 10 and a display device according to claim 11, said server being communicatively connected to said display device and said terminal device, respectively.

13. A storage medium for computer-readable storage, characterized in that the storage medium stores one or more programs which are executable by one or more processors to implement the steps of the method of AI model deployment of any of claims 1 to 6 or claims 7 to 9.