Exhibition hall interactive display method and system supporting artificial intelligence recognition

Abstract

The invention belongs to the technical field of artificial intelligence, and discloses an exhibition hall interactive display method and system supporting artificial intelligence recognition, and the method comprises the steps: collecting the multi-modal behavior information of audiences in an exhibition hall environment, recognizing the interaction intention of the audiences through behavior pattern analysis and semantic understanding, and displaying the interaction intention of the audiences. Mapping the interaction intention of the audience to an exhibition hall virtual scene to form a semantic interaction node space; according to the interaction intention of the audience, the association relationship between each node in the semantic interaction node space and the exhibit or exhibition hall scene element is analyzed, the display content capable of being triggered is determined, and a corresponding interaction display request set is generated; when different audiences trigger the interactive display requests at the same time, coordinating the interactive display request set, and adjusting and scheduling each interactive display request according to exhibition hall resource limitation and a preset priority rule; interaction intelligence and self-adaption capability of the exhibition hall are enhanced, and fluency and immersion of audience interaction experience are improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of deep learning, more particularly, to an exhibition hall interactive display method and system supporting artificial intelligence recognition. BACKGROUND

[0002] The existing exhibition hall interactive display method and system supporting artificial intelligence recognition mainly has the following problems: With the development of artificial intelligence and immersive display technology, exhibition hall interaction display gradually evolves from static one-way presentation to dynamic, multi-modal human-computer interaction. Although the existing exhibition hall interactive display method and system can trigger display content based on the actions of the audience, it still has many shortcomings.

[0003] The existing technology generally adopts a single action directly mapping display content mode, lacking semantic association modeling between audience intent and display content. Different audiences often trigger the same display content, which cannot be dynamically optimized according to the audience's historical interaction records, behavior patterns and personalized needs, resulting in a mismatch between the display results and the audience's real needs, making it difficult to provide intelligent and immersive interactive experience.

[0004] When there are multiple audiences in the exhibition hall triggering interactive requests at the same time, the existing system usually adopts a queuing mechanism or a fixed priority table for processing, such as first come first served or pre-set priority sorting. Although this approach can avoid device resource conflicts, it is difficult to consider the matching degree of audience interaction intent and display content, and it also fails to consider the limited resource constraints of the exhibition hall. Therefore, it is easy to cause partial resource vacancy or overload, lack of dynamic scheduling and fair optimization mechanism for multiple audience requests, resulting in insufficient overall exhibition experience.

[0005] In the rendering execution process, the existing exhibition hall interactive display method usually only considers the logical trigger of the display content, but lacks real-time monitoring of the resource occupation of the rendering request in the display terminal, computing device and rendering channel. The existing method is mostly based on fixed priority or round-robin strategy for scheduling, when concurrent rendering requests occur, the overall resource utilization of the system is easy to exceed the threshold, causing frame dropping, lagging or even display interruption. In the case of resource overload, the existing technology lacks a unified dynamic adjustment mechanism, often simply discarding some audience requests, which not only affects the fairness of interaction, but also seriously reduces the audience's immersive experience.

[0006] In view of this, the present application proposes an exhibition hall interactive display method supporting artificial intelligence recognition to solve the above problems. SUMMARY

[0007] In order to overcome the above-mentioned defects of the prior art, in order to achieve the above-mentioned purpose, the present application provides the following technical scheme: an exhibition hall interactive display method supporting artificial intelligence recognition, comprising: S1, collect the multi-modal behavior information of the audience in the exhibition hall environment, identify the interactive intention of the audience through behavior pattern analysis and semantic understanding, and map the interactive intention of the audience to the virtual scene of the exhibition hall to form a semantic interaction node space; S2, according to the interactive intention of the audience, analyze the association relationship between each node in the semantic interaction node space and the exhibition or exhibition hall scene element, determine the triggerable display content, and generate a corresponding interactive display request set; S3, when there are different audiences triggering interactive display requests at the same time, coordinate the interactive display request set, adjust and schedule each interactive display request according to the exhibition hall resource limit and the preset priority rule; S4, convert the adjusted and scheduled interactive display request into a corresponding rendering request, and submit the rendering request to the exhibition hall interaction terminal. Monitor the resource amount occupied by each rendering request in the rendering process, obtain the exhibition hall resource utilization rate, generate a rendering strategy based on the exhibition hall resource utilization rate, and dynamically display the exhibition hall content according to the rendering strategy.

[0008] Specifically, the multi-modal behavior information includes visual modal information, auditory modal information, motion modal information and environmental interaction modal information.

[0009] Specifically, the method for identifying the interactive intention of the audience includes: The acquired multi-modal behavior information is denoised, and a multi-modal feature vector is extracted. The multi-modal feature vector includes visual modal features, auditory modal features, motion modal features and environmental interaction modal features. The multi-modal feature vector is input into a pre-constructed behavior recognition model, and the behavior recognition model outputs an action category label and an emotion state label of the audience. The action category label and the emotion state label are matched in combination with a preset rule library to form a corresponding behavior pattern label. The behavior recognition model is any one of a convolutional neural network model or a recurrent neural network model. The behavior pattern label is semantically matched with the exhibition hall scene element based on a knowledge graph, and the interactive intention of the audience is inferred.

[0010] Specifically, the method for forming the semantic interaction node space includes: A virtual scene is constructed in the exhibition hall environment, a corresponding virtual model is established for each exhibit, exhibition area and interaction device, and scene semantic labels are attached to the virtual model. A semantic annotation relationship table is established, the identified interactive intention of the audience is analyzed, and the core elements of the interactive intention are extracted, including intention type, target object, interaction method and trigger time. The interactive intention is bound to a scene element in the virtual model through semantic matching to generate an interaction node; after the interactive intentions of different audiences are identified and the interaction nodes are generated, the interaction nodes are organized according to time sequence and spatial distribution to form a semantic interaction node space.

[0011] Specifically, the method for generating the corresponding interactive display request comprises: A display hall display content set is established, each display content corresponds to multimedia explanation of an exhibit, immersive display of an exhibition area or experience program of an interactive device, and a content semantic label and a relevance weight are attached to the display content; Context information of the audience is collected, the identified audience interactive intention is structured and expressed, the context information of the audience is combined, and the probability value of the audience interactive intention belonging to each content semantic label is estimated through Bayesian inference; The probability value and the relevance weight of the display content are comprehensively calculated to obtain the matching score between the audience interactive intention and each display content; when the matching score exceeds a preset matching score threshold, the corresponding display content is determined as a triggerable content, and all display contents meeting the condition are aggregated into an interactive display request set.

[0012] Specifically, the method for adjusting and scheduling each interactive display request comprises: When different audiences trigger interactions at the same time, the interactive display request set is coordinated, a total amount of available resources of the exhibition hall is preset, the total amount of available resources is all schedulable resources, the schedulable resources include projection equipment, rendering computing power, display screens and interactive devices; for each display content, a required resource consumption amount in the display process is preset; The exhibition hall resource limitation and the preset priority rule need to be met at the same time in the scheduling process, the exhibition hall resource limitation includes that the total amount of resource consumption required by all allocated display contents cannot exceed the total amount of available resources of the exhibition hall, and each audience can be allocated at most one display content; When the exhibition hall resource limitation is met, when different audiences trigger interactive requests at the same time, the interactive requests of different audiences are sorted according to the preset priority rule, a distribution variable is set for each combination of audience and display content, which is used to represent whether the audience is allocated to the display content; when the distribution is established, the distribution variable takes a value of 1, otherwise, the value is 0, so as to dynamically adjust the allocation result of each interactive display request.

[0013] Specifically, the method for converting the adjusted and scheduled interactive display request into a corresponding rendering request comprises: The adjusted interactive display request is parsed to obtain display content, scene position, presentation form and interaction parameters corresponding to each interactive display request; the display content corresponding to each interactive display request is mapped to a rendering resource, and the spatial layout of the display content in the virtual scene is determined according to the scene position, and the display mode is determined according to the presentation form; according to the rendering engine interface specification, the display content, scene position, presentation form and interaction parameters are converted into rendering instructions; the rendering instructions are encapsulated to form an executable rendering request.

[0014] Specifically, the method for obtaining the resource utilization rate of the exhibition hall comprises: The rendering request is submitted to the exhibition hall interactive terminal, and the resource amount occupied by each rendering request, including the use of display terminals, computing devices and rendering channels, is monitored in real time during the rendering process; the resource occupation amount of each rendering request is recorded and associated with the corresponding audience and display content to form resource usage state information; the resource occupation amount of each rendering request is associated with the allocation state of the rendering request, and the resource utilization rate of the exhibition hall is calculated and obtained.

[0015] Specifically, the method for dynamically displaying the content of the exhibition hall according to the rendering strategy comprises: A preset exhibition hall resource utilization rate threshold is set, when the resource utilization rate of the exhibition hall is less than or equal to the preset exhibition hall resource utilization rate threshold, all rendering requests are allowed to be rendered according to the allocation variable at time point t; when the resource utilization rate of the exhibition hall is greater than the preset exhibition hall resource utilization rate threshold, a dynamic allocation adjustment mechanism is used to adjust the allocation state of the rendering request, and a corresponding rendering strategy is generated, and the content of the exhibition hall is dynamically displayed according to the rendering strategy.

[0016] An exhibition hall interactive display system supporting artificial intelligence recognition, comprising: An audience behavior recognition unit collects multi-modal behavior information of audiences in the exhibition hall environment, recognizes the interactive intention of the audiences through behavior pattern analysis and semantic understanding, and maps the interactive intention of the audiences to the virtual scene of the exhibition hall to form a semantic interaction node space; A display content generation unit analyzes the association relationship between each node in the semantic interaction node space and the exhibits or exhibition hall scene elements according to the interactive intention of the audiences, determines the display content that can be triggered, and generates a corresponding interactive display request set; A display content scheduling unit coordinates the interactive display request set when different audiences trigger the interactive display request at the same time, adjusts and schedules each interactive display request according to the exhibition hall resource limitation and the preset priority rule; The interactive content display unit converts the adjusted interactive display request into a corresponding rendering request, submits the rendering request to the exhibition hall interactive terminal, monitors the resource amount occupied by each rendering request in the rendering process, obtains the exhibition hall resource utilization, generates a rendering strategy based on the exhibition hall resource utilization, and dynamically displays the exhibition hall content according to the rendering strategy.

[0017] Compared with the prior art, the present application has the following beneficial effects: The present application estimates the intention probability value by combining Bayesian inference with context information, overcomes the recognition bias under a single trigger mode, can more finely distinguish the true intentions of different audiences, and comprehensively considers the audience context information and the semantic weight of the display content through the matching score, so that different audiences can trigger display content that meets their needs, improving the differentiation and immersion of the interactive experience. By monitoring the resource occupation of each rendering request in the rendering process in real time, the exhibition hall resource utilization is calculated in combination with the allocation state, thereby establishing a feedback regulation mechanism that can dynamically perceive the load of the exhibition hall, accurately reflecting the real-time resource consumption of the system in terms of display terminals, computing devices and rendering channels. The proposed dynamic allocation regulation mechanism can proportionally reduce the allocation state of the rendering request, rather than simply discarding low-priority requests. It ensures that all rendering requests obtain the corresponding actual resource quota in the case of resource shortage, realizes the fairness and continuity of resource allocation. Based on the exhibition hall resource utilization, the corresponding rendering strategy is generated, which supports normal rendering of all requests when the resources are sufficient, and maintains the overall utilization rate below the threshold through dynamic adjustment when the resources are scarce, thereby guaranteeing the continuous presentation of the display content and the stability of the system operation, and improving the immersive experience of the audience. BRIEF DESCRIPTION OF DRAWINGS

[0018] Figure 1 It is a flowchart of an exhibition hall interactive display method supporting artificial intelligence recognition. Figure 2 It is a structural diagram of an exhibition hall interactive display system supporting artificial intelligence recognition. DETAILED DESCRIPTION

[0019] With reference to the accompanying drawings, the technical solutions in the embodiments of the present application will be clearly and completely described below, obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present application. Embodiments

[0020] Please refer to Figure 1 The present embodiment provides an exhibition hall interactive display method supporting artificial intelligence recognition, and specifically comprises the following steps: S1, multi-modal behavior information of visitors in the exhibition hall environment is collected, the interactive intention of the visitors is recognized through behavior pattern analysis and semantic understanding, and the interactive intention of the visitors is mapped to the virtual scene of the exhibition hall to form a semantic interaction node space; S2, according to the interactive intention of the visitors, the association between each node in the semantic interaction node space and the exhibit or the exhibition hall scene element is analyzed, the display content that can be triggered is determined, and a corresponding interactive display request set is generated; S3, when different visitors trigger interactive display requests at the same time, the interactive display request set is coordinated, each interactive display request is adjusted and scheduled according to the exhibition hall resource limit and the preset priority rule; S4, the adjusted and scheduled interactive display request is converted into a corresponding rendering request, and the rendering request is submitted to the exhibition hall interactive terminal. The resource amount occupied by each rendering request is monitored in the rendering process, the exhibition hall resource utilization rate is obtained, the rendering strategy is generated based on the exhibition hall resource utilization rate, and the exhibition hall content is dynamically displayed according to the rendering strategy.

[0021] The multi-modal behavior information includes visual modal information, auditory modal information, action modal information and environmental interaction modal information.

[0022] The visual modal information includes posture information, gesture information, expression information and gaze direction information of the visitors; the auditory modal information includes voice content information, voice tone information and sound source positioning information of the visitors; the action modal information includes position trajectory information of the visitors in the exhibition hall and distance information between the visitors and the exhibits; the environmental interaction modal information includes interactive terminal operation information and visitor stay time; the interactive terminal operation information includes touch screen clicking, sliding, mobile terminal scanning and APP control.

[0023] The method for recognizing the interactive intention of the visitors includes: The obtained multi-modal behavior information is denoised, and a multi-modal feature vector is extracted; the multi-modal feature vector contains visual modal features, auditory modal features, action modal features and environmental interaction modal features; The specific acquisition approach of the multi-modal feature vector is: extracting the posture key points, gesture type, facial expression feature vector and gaze direction coordinates of the audience from the visual modal information to form the visual modal feature; extracting the speech keyword, tone curve and sound source positioning information from the auditory modal information to form the auditory modal feature; extracting the moving path, stay area and spatial distance change information of the audience from the action modal information to form the action modal feature; extracting the interactive operation type and stay duration from the environmental interaction modal information to form the environmental interaction modal feature; inputting the multi-modal feature vector into a pre-constructed behavior recognition model, the behavior recognition model outputting the action category label and the emotional state label of the audience; matching the action category label and the emotional state label in combination with a pre-set rule base to form a corresponding behavior mode label; The behavior recognition model is any one of a convolutional neural network model or a recurrent neural network model; the behavior mode label is semantically matched with the exhibition hall scene elements based on a knowledge graph to infer the interactive intention of the audience.

[0024] For example: the audience stays in front of a certain cultural relic showcase, points to the showcase and says "Can this be enlarged?", the system first identifies the gesture type and gaze direction coordinates through the visual modal information, identifies the speech keyword "enlarge" in combination with the auditory modal information, obtains the distance between the audience and the exhibit through the action modal information, and finally infers and determines that the interactive intention of the audience is "requesting the exhibit to be enlarged for display" in combination with the pre-set rule base, and generates a corresponding interactive node to trigger the exhibit enlargement demonstration function in the virtual scene.

[0025] The method for forming the semantic interactive node space comprises: building a virtual scene in the exhibition hall environment, establishing corresponding virtual models for each exhibit, exhibition area and interactive device, and attaching scene semantic labels such as exhibit ID, exhibition area function, interactive mode, etc. to the virtual models, and simultaneously establishing a semantic annotation relationship table for supporting subsequent intention mapping operations; the virtual model is a digital twin model; analyzing the recognized interactive intention of the audience, extracting the core elements of the interactive intention, the core elements including the intention type, target object, interactive mode and trigger time; wherein the intention type can be explanation, enlargement, comparison or navigation, etc., and the target object corresponds to the exhibit or functional element in the virtual scene; binding the interactive intention and the scene elements in the virtual model through semantic matching to generate an interactive node, the interactive node containing node ID, intention type, trigger audience identity, target element, trigger timestamp and intention confidence value, etc.; the interactive node is a logical trigger point in the virtual scene and is used to drive the subsequent display content; After the interactive intentions of different audiences are identified and the interactive nodes are generated, the interactive nodes are organized according to time sequence and spatial distribution to form a semantic interactive node space. The interactive nodes are connected through relationship edges, and the relationship edges are used to represent "continuous interactive intentions of the same audience", "interactive intentions of different audiences for the same exhibit" or "interactive intention migration across exhibition areas"; For example, when a certain audience says "Can this be enlarged?" in front of a showcase, the system identifies the interactive intention as "requesting enlarged display of the exhibit", and finds the virtual model node of the corresponding exhibit in the virtual scene. The system generates an interactive node, records the information as: intention type: enlarged display, audience ID: 001, target exhibit: showcase A-antique X, timestamp: 2025-06-21, and adds the interactive node to the semantic interactive node space. At the same time, the interactive node and the "request for explanation" interactive node of the same audience are connected through a relationship edge, which is used to depict the continuous interactive intention trajectory of the same audience. Finally, the display scheduling module triggers the enlarged demonstration function according to the interactive node to realize the interaction between the audience and the exhibition hall scene.

[0026] The method for generating the corresponding interactive display request comprises: establishing a set of exhibition hall display contents, each display content corresponding to a multimedia explanation of an exhibit, an immersive display of an exhibition area or an experience program of an interactive device, and adding content semantic labels and relevance weights to the display contents; collecting the context information of the audience, structurally expressing the identified interactive intention of the audience, combining the context information of the audience, and estimating the probability value of the interactive intention of the audience belonging to each content semantic label through Bayesian inference; the context information of the audience includes the position of the audience, the current behavior mode (the body posture of the audience, such as standing, moving and staying) and the historical interactive record; comprehensively calculating the probability value and the relevance weight of the display content to obtain the matching score between the interactive intention of the audience and each display content; when the matching score exceeds a preset matching score threshold, the corresponding display content is determined as a triggerable content, and all display contents meeting the condition are aggregated into an interactive display request set.

[0027] The matching score is: ; wherein, represents the matching score of the audience and the display content ; represents the set of all content semantic labels; represents one of the content semantic labels in the set of content semantic labels; represents the display content and the content semantic label correlation weight of the relatedness, an initial weight value given according to the audience historical interaction data, for example, the weight of the "exhibition introduction video" and the "query information" intent is marked as 0.9, and the weight of the "trigger interactive experience" intent is marked as 0.2; represents the context information of the given audience under the condition that the audience interactive intent belongs to the content semantic label probability value; represents the interactive intent variable of the audience ; represents the context information of the audience; Solves the following technical problems existing in the prior art: the prior art often directly maps a single action to a display content, resulting in a mismatch between interactive content and audience needs; lacks modeling of the semantic association between intent and content, different audiences may trigger the same content, and lacks personalization; the prior art fails to incorporate audience historical interaction records, behavior patterns and other information into the inference process, and cannot dynamically optimize content matching; lacks an interactive content selection mechanism based on probability inference, making it difficult for the display system to achieve a more intelligent and immersive interactive experience.

[0028] The beneficial effects of the prior art are: through Bayesian inference combined with context information to estimate the intent probability value, the recognition bias under a single trigger mode is overcome, and different audiences' real intents can be more finely distinguished; by matching score, the audience context information and the semantic weight of the display content are considered comprehensively, so that different audiences can trigger display content that meets their needs, improving the differentiation and immersion of the interactive experience; the weight relationship between the display content and the intent semantic label establishes a bridge at the semantic level, and more reasonable content scheduling can be achieved based on the semantic level; by setting a matching score threshold, the triggering of the display content can be dynamically determined according to the audience behavior and context information, enhancing the intelligence and flexibility of the interactive display; The method for adjusting and scheduling each interactive display request comprises: When different audiences in the exhibition hall trigger interactive requests at the same time, the prior art usually only uses a simple queuing mechanism or priority rule, such as processing the display requests in turn according to the first-come-first-served principle or a fixed priority table. Although this method can avoid conflicts of exhibition hall device resources, it has the following disadvantages: it cannot comprehensively consider the matching degree of audience interactive intent and display content, and may easily result in a mismatch between display content and audience needs; ignores the limited resource constraints of the exhibition hall (such as projection device bandwidth, rendering computing power, and the number of display screens), which may result in partial resource idling or overload, lacks a dynamic optimization mechanism, and is difficult to meet the fairness and overall viewing experience of multiple audience interactive requests at the same time.

[0029] To solve the above problems, when different audiences trigger interactions simultaneously, the interactive display request set is coordinated, the total amount of available resources of the exhibition hall is preset, and the total amount of available resources includes all resources available for scheduling, including projection equipment, rendering computing power, display screens, and interactive devices. For each display content, the resource consumption required during the display process is preset. In the scheduling process, the exhibition hall resource limit and the preset priority rule need to be met simultaneously. The exhibition hall resource limit includes that the total resource consumption required by all allocated display contents cannot exceed the total amount of available resources of the exhibition hall, and each audience can be allocated at most one display content. When different audiences trigger interactive requests simultaneously while meeting the exhibition hall resource limit, the interactive requests of different audiences are sorted according to the preset priority rule. The preset priority rule includes sorting the interactive requests of different audiences according to the matching score between the audience's interactive intention and each display content. For each combination of an audience and a display content, an allocation variable is set to indicate whether the audience is allocated to the display content. The allocation variable is: When the allocation is established, the allocation variable takes the value 1, otherwise it takes the value 0, thereby dynamically adjusting the allocation results of each interactive display request.

[0030] The method for converting the adjusted interactive display request into a corresponding rendering request includes: The adjusted interactive display request is parsed to obtain the display content, scene position, presentation form, and interaction parameters corresponding to each interactive display request. The display content corresponding to each interactive display request is mapped to a rendering resource, and the spatial layout of the display content in the virtual scene is determined according to the scene position, and the display method (such as two-dimensional, three-dimensional, animation, or multimedia effect) is determined according to the presentation form. According to the rendering engine interface specification, the display content, scene position, presentation form, and interaction parameters are converted into rendering instructions, including resource loading instructions, material binding instructions, animation setting instructions, lighting adjustment instructions, camera view angle adjustment instructions, and event binding instructions. The rendering instructions are encapsulated to form executable rendering requests, so that the exhibition hall interactive terminal can directly read and execute the corresponding rendering instructions.

[0031] The method for obtaining the utilization rate of exhibition hall resources includes: The rendering request is submitted to the exhibition hall interactive terminal, and the resource usage of each rendering request, including the display terminal, computing device, and rendering channel, is monitored in real time during the rendering process. The resource usage of each rendering request is recorded and associated with the corresponding audience and display content to form resource usage state information. The resource usage of each rendering request is associated with the allocation state of the rendering request, and the utilization rate of the exhibition hall resources is calculated and obtained.

[0032] the utilization rate of the exhibition hall resources is: ; wherein, represents the utilization rate of the exhibition hall resources at time point ; represents the resource occupation of the corresponding rendering request of the display content ; represents the allocation variable at time point , that is, whether the display content is allocated to the audience at time point ; represents the total amount of available resources of the preset exhibition hall; The method for dynamically displaying the exhibition hall content according to the rendering strategy comprises: a preset utilization rate threshold of the exhibition hall resources, when the utilization rate of the exhibition hall resources is less than or equal to the preset utilization rate threshold of the exhibition hall resources, all rendering requests are allowed to be rendered according to the allocation variable at time point t; when the utilization rate of the exhibition hall resources is greater than the preset utilization rate threshold of the exhibition hall resources, a dynamic allocation adjustment mechanism is used to adjust the allocation state of the rendering request, a corresponding rendering strategy is generated, and the exhibition hall content is dynamically displayed according to the rendering strategy.

[0033] The dynamic allocation adjustment mechanism is: ; wherein, represents the adjusted allocation variable at time point , representing the actual resource quota of each rendering request; represents the preset utilization rate threshold of the exhibition hall resources; represents a scaling factor, when the utilization rate of the exhibition hall resources is greater than the preset utilization rate threshold of the exhibition hall resources, the resource allocation of all rendering requests needs to be reduced in proportion to ensure that the reduced utilization rate of the exhibition hall resources is pressed back below the preset utilization rate threshold of the exhibition hall resources.

[0034] The following technical problems existing in the prior art are solved: In the rendering process of the existing exhibition hall interactive display, only the logical trigger of the display content is considered, and real-time monitoring of the resource occupation of the rendering request in the display terminal, the computing device and the rendering channel is lacking, which easily leads to uneven resource allocation, and some display content is stuck or delayed. The existing method generally performs rendering resource scheduling based on a fixed priority or a polling strategy, which leads to that when the exhibition hall appears concurrent rendering requests, the overall resource utilization rate of the system exceeds the threshold, causing frame dropping or display interruption. When the resources are overloaded, there is a lack of unified dynamic adjustment mechanism, which cannot maintain the overall performance while ensuring fairness, and some audience's requests are easily discarded completely, resulting in poor user experience.

[0035] ​Compared to existing technologies, the advantages are as follows: By monitoring the resource consumption of each rendering request in real time during the rendering process and calculating the exhibition hall resource utilization rate based on the allocation status, a feedback adjustment mechanism capable of dynamically sensing the exhibition hall load is established, accurately reflecting the real-time resource consumption of the system in terms of display terminals, computing devices, and rendering channels. The proposed dynamic allocation adjustment mechanism can proportionally reduce and adjust the allocation status of rendering requests, rather than simply discarding low-priority requests. This ensures that all rendering requests receive the corresponding actual resource quota under resource constraints, achieving fairness and continuity in resource allocation. Based on the exhibition hall resource utilization rate, a corresponding rendering strategy is generated. When resources are sufficient, all requests are supported for normal rendering. When resources are scarce, dynamic adjustment keeps the overall utilization rate below a threshold, thereby ensuring the continuous presentation of the displayed content and the stability of system operation, and enhancing the immersive experience for the audience.

[0036] The preset threshold for exhibition hall resource utilization is set by staff based on historical data analysis results. This historical analysis process includes the system collecting the resource utilization rates of multiple exhibition halls and calculating their average value as a reference to obtain the preset threshold for exhibition hall resource utilization. This threshold can be adjusted by staff during system operation according to the actual situation.

[0037] This embodiment estimates the probability value of intent by combining Bayesian inference with contextual information, overcoming the recognition bias in the single-trigger mode and enabling more precise differentiation of the true intent of different viewers. By comprehensively considering the audience's contextual information and the semantic weight of the displayed content through matching scores, different viewers can trigger display content that meets their needs, enhancing the differentiation and immersion of the interactive experience. The weight relationship between the displayed content and the semantic tags of intent establishes a bridge at the semantic layer, enabling more reasonable content scheduling based on the semantic level. By setting a matching score threshold, the triggering of displayed content can be dynamically determined based on audience behavior and contextual information, enhancing the intelligence and flexibility of the interactive display. By monitoring the resource consumption of each rendering request in real time during the rendering process and calculating the exhibition hall resource utilization rate based on the allocation status, a feedback adjustment mechanism capable of dynamically sensing the exhibition hall load was established. This accurately reflects the real-time resource consumption of the system in terms of display terminals, computing devices, and rendering channels. The proposed dynamic allocation adjustment mechanism can proportionally reduce and adjust the allocation status of rendering requests, rather than simply discarding low-priority requests. This ensures that all rendering requests receive the corresponding actual resource quota under resource constraints, achieving fairness and continuity in resource allocation. Based on the exhibition hall resource utilization rate, a corresponding rendering strategy is generated. When resources are sufficient, all requests are supported for normal rendering. When resources are scarce, dynamic adjustment keeps the overall utilization rate below a threshold, thereby ensuring the continuous presentation of the displayed content and the stability of system operation, and enhancing the immersive experience for visitors. Example

[0038] Please refer to Figure 2 As shown in the embodiment, some of the descriptions are provided in the embodiment 1, and the embodiment provides an exhibition hall interactive display system supporting artificial intelligence recognition, including: An audience behavior recognition unit collects multi-modal behavior information of the audience in the exhibition hall environment, recognizes the interactive intention of the audience through behavior pattern analysis and semantic understanding, and maps the interactive intention of the audience to the virtual scene of the exhibition hall to form a semantic interaction node space. A display content generation unit analyzes the association between each node in the semantic interaction node space and the exhibit or the exhibition hall scene element according to the interactive intention of the audience, determines the display content that can be triggered, and generates a corresponding interactive display request set. A display content scheduling unit coordinates the interactive display request set when different audiences trigger the interactive display request at the same time, adjusts and schedules each interactive display request according to the exhibition hall resource limit and the preset priority rule. An interactive content display unit converts the adjusted and scheduled interactive display request into a corresponding rendering request, submits the rendering request to the exhibition hall interactive terminal, monitors the resource amount occupied by each rendering request in the rendering process, obtains the exhibition hall resource utilization rate, generates a rendering strategy based on the exhibition hall resource utilization rate, and dynamically displays the exhibition hall content according to the rendering strategy.

[0039] Since the electronic device introduced in the embodiment is the electronic device used in the implementation of the exhibition hall interactive display method and system supporting artificial intelligence recognition in the embodiment, the specific implementation of the electronic device and its various forms can be understood by those skilled in the art based on the exhibition hall interactive display method and system supporting artificial intelligence recognition introduced in the embodiment. Therefore, how the electronic device implements the method in the embodiment will not be described in detail. As long as the electronic device used in the implementation of the exhibition hall interactive display method and system supporting artificial intelligence recognition in the embodiment is implemented by those skilled in the art, it belongs to the scope of protection of the present application.

[0040] The above formulas are dimensionless numerical calculations, and the formulas are obtained by software simulation of a large amount of data to obtain a formula of the most recent real situation. The preset parameters and threshold values in the formula are set by those skilled in the art according to the actual situation.

[0041] The above is only the preferred embodiment of the present application, and the protection scope of the present application is not limited to the above-mentioned embodiments. Any technical solution that belongs to the idea of the present application is within the protection scope of the present application. It should be noted that some improvements and decorations that do not deviate from the principle of the present application can also be considered as the protection scope of the present application.

Claims

1. An exhibition hall interactive display method supporting artificial intelligence recognition, characterized in that, The method comprises the following steps: S1, collecting multi-modal behavior information of visitors in the exhibition hall environment, identifying the interactive intention of the visitors through behavior pattern analysis and semantic understanding, and mapping the interactive intention of the visitors to the virtual scene of the exhibition hall to form a semantic interaction node space; S2, according to the interactive intention of the visitors, analyzing the association between each node in the semantic interaction node space and the exhibition or exhibition hall scene elements, determining the triggerable display content, and generating a corresponding interactive display request set; S3, when there are different visitors triggering interactive display requests at the same time, coordinating the interactive display request set, adjusting and scheduling each interactive display request according to the exhibition hall resource limit and the preset priority rule; S4, converting the adjusted and scheduled interactive display request into a corresponding rendering request, submitting the rendering request to the exhibition hall interaction terminal, monitoring the resource amount occupied by each rendering request in the rendering process, obtaining the exhibition hall resource utilization rate, generating a rendering strategy based on the exhibition hall resource utilization rate, and dynamically displaying the exhibition hall content according to the rendering strategy.

2. The exhibition hall interactive display method supporting artificial intelligence recognition according to claim 1, characterized in that, The multi-modal behavior information includes visual modal information, auditory modal information, motion modal information and environmental interaction modal information.

3. The exhibition hall interactive display method supporting artificial intelligence recognition according to claim 2, characterized in that, The method for identifying the interactive intention of the visitors comprises: de-noising the obtained multi-modal behavior information and extracting a multi-modal feature vector; the multi-modal feature vector includes visual modal features, auditory modal features, motion modal features and environmental interaction modal features; inputting the multi-modal feature vector into a pre-constructed behavior recognition model, and the behavior recognition model outputs action category labels and emotion state labels of the visitors; combining a preset rule base to match the action category labels and the emotion state labels to form corresponding behavior pattern labels; the behavior recognition model is any one of a convolutional neural network model or a recurrent neural network model; the behavior pattern labels are semantically matched with the exhibition hall scene elements based on a knowledge graph to infer the interactive intention of the visitors.

4. The exhibition hall interactive display method supporting artificial intelligence recognition according to claim 3, characterized in that, The method for forming the semantic interaction node space comprises: building a virtual scene in the exhibition hall environment, establishing corresponding virtual models for each exhibit, exhibition area and interactive device, and attaching scene semantic labels to the virtual models, while establishing a semantic annotation relationship table, analyzing the identified interactive intention of the visitors, and extracting core elements of the interactive intention, including intention type, target object, interaction mode and trigger time; binding the interactive intention with the scene elements in the virtual models through semantic matching to generate interaction nodes; after the interactive intention of different visitors is identified and the interaction nodes are generated, the interaction nodes are organized according to time sequence and spatial distribution to form a semantic interaction node space.

5. The exhibition hall interactive display method supporting artificial intelligence recognition according to claim 4, characterized in that, The method for generating corresponding interactive display requests comprises: establishing an exhibition hall display content set, each display content corresponding to multimedia explanation of an exhibit, immersive display of an exhibition area or experience program of an interactive device, and attaching content semantic labels and relevance weights to the display content; collecting context information of the visitors, structurally expressing the identified interactive intention of the visitors, combining the context information of the visitors, and estimating the probability value of the interactive intention of the visitors belonging to each content semantic label through Bayesian inference. The probability value is combined with the relevance weight of the display content to obtain a matching score between the audience interaction intention and each display content; when the matching score exceeds a preset matching score threshold, the corresponding display content is determined as triggerable content, and all display contents meeting the condition are aggregated into an interactive display request set.

6. The exhibition hall interactive display method supporting artificial intelligence identification according to claim 5, characterized in that, The method for adjusting and scheduling each interactive display request comprises: When different audiences trigger interactions at the same time, the interactive display request set is coordinated, the total amount of available resources of the preset exhibition hall is determined, the total amount of available resources comprises all schedulable resources, the schedulable resources comprise projection equipment, rendering computing power, display screens and interactive devices; for each display content, the required resource consumption amount in the display process is preset; In the scheduling process, the exhibition hall resource limitation and the preset priority rule need to be met at the same time, the exhibition hall resource limitation comprises that the total amount of resource consumption required by all allocated display contents cannot exceed the total amount of available resources of the exhibition hall, and each audience can be allocated at most one display content; When the exhibition hall resource limitation is met, when different audiences trigger interactive requests at the same time, the interactive requests of different audiences are sorted according to the preset priority rule, an allocation variable is set for each combination of audience and display content, which is used to represent whether the audience is allocated to the display content; when the allocation is established, the allocation variable takes the value 1, otherwise it takes the value 0, thereby dynamically adjusting the allocation result of each interactive display request.

7. The exhibition hall interactive display method supporting artificial intelligence recognition according to claim 6, characterized in that, The method for converting the adjusted and scheduled interactive display request into a corresponding rendering request comprises: The adjusted and scheduled interactive display request is analyzed to obtain the display content, scene position, presentation form and interaction parameter corresponding to each interactive display request; the display content corresponding to each interactive display request is mapped to a rendering resource, and the spatial layout of the display content in the virtual scene is determined according to the scene position, and the display mode is determined according to the presentation form; according to the interface specification of the rendering engine, the display content, scene position, presentation form and interaction parameter are converted into rendering instructions, the rendering instructions are encapsulated to form executable rendering requests.

8. The exhibition hall interactive display method supporting artificial intelligence recognition according to claim 7, characterized in that, The method for obtaining the exhibition hall resource utilization rate comprises: The rendering request is submitted to the exhibition hall interactive terminal, the resource amount occupied by each rendering request is monitored in real time in the rendering process, including the use of resources of the display terminal, computing device and rendering channel; the resource occupation amount of each rendering request is recorded and associated with the corresponding audience and display content to form resource use state information; the resource occupation amount of each rendering request is associated with the allocation state of the rendering request, and the exhibition hall resource utilization rate is calculated and obtained.

9. The exhibition hall interactive display method supporting artificial intelligence recognition according to claim 8, characterized in that, The method for dynamically displaying the exhibition hall content according to the rendering strategy comprises: A preset exhibition hall resource utilization rate threshold is set, when the exhibition hall resource utilization rate is less than or equal to the preset exhibition hall resource utilization rate threshold, all rendering requests are allowed to be rendered according to the allocation variable at time point t; when the exhibition hall resource utilization rate is greater than the preset exhibition hall resource utilization rate threshold, a dynamic allocation adjustment mechanism is adopted to adjust the allocation state of the rendering request, a corresponding rendering strategy is generated, and the exhibition hall content is dynamically displayed according to the rendering strategy.

10. An exhibition hall interactive display system supporting artificial intelligence recognition, used to implement the exhibition hall interactive display method supporting artificial intelligence recognition in any one of claims 1 to 9, characterized in that, Comprise: The audience behavior recognition unit collects multi-modal behavior information of the audience in the exhibition hall environment, recognizes the interactive intention of the audience through behavior pattern analysis and semantic understanding, and maps the interactive intention of the audience to the virtual scene of the exhibition hall to form a semantic interaction node space. The display content generation unit analyzes the association between each node in the semantic interaction node space and the exhibit or the exhibition hall scene element according to the interactive intention of the audience, determines the display content that can be triggered, and generates a corresponding interactive display request set. The display content scheduling unit coordinates the interactive display request set when different audiences simultaneously trigger the interactive display request, adjusts and schedules each interactive display request according to the exhibition hall resource limit and the preset priority rule. The interactive content display unit converts the adjusted and scheduled interactive display request into a corresponding rendering request, submits the rendering request to the exhibition hall interaction terminal, monitors the resource amount occupied by each rendering request in the rendering process, obtains the exhibition hall resource utilization rate, generates a rendering strategy based on the exhibition hall resource utilization rate, and dynamically displays the exhibition hall content according to the rendering strategy.

Images