A Food Video Display and Interactive System and Method Based on User Feedback

Abstract

This invention relates to an interactive food video display system and method based on user feedback. By integrating visitor behavior records and audio signals to form comprehensive feedback data, it achieves dynamic updates of video content and cyclical optimization of playlists. This addresses the shortcomings of traditional food video display methods, significantly enhancing the interactivity and appeal of exhibitions. This method not only strengthens the interactivity of exhibitions, enabling the displayed content to respond to visitors' interests and reactions in real time, but also effectively extends visitors' dwell time, improving their visitor experience quality and satisfaction. By continuously optimizing the displayed content, it ensures that each visit brings visitors a fresh and personalized experience, significantly enhancing the overall appeal and interactivity of exhibitions.

Description

Technical Field

[0001] This invention belongs to the field of interactive display systems, specifically relating to a food video display and interactive system and method based on user feedback. Background Technology

[0002] In the current exhibition field, especially in food-themed exhibitions, a common practice is to pre-set a series of fixed video clips that are played on a loop across multiple screens within the exhibition hall. These clips typically include elaborate food preparation processes, enticing finished dishes, and related introductions to food culture. However, this traditional approach suffers from a significant technical problem: a lack of responsiveness to real-time visitor feedback. Because the video content remains unchanged, it cannot be dynamically adjusted based on visitor interests and behaviors, resulting in a rather monotonous experience and failing to effectively attract and maintain visitor attention.

[0003] Existing methods for showcasing food through videos primarily rely on pre-planned content, neglecting real-time feedback and personalized needs from visitors. This not only limits the interactivity of exhibitions but also fails to fully utilize modern sensing technologies and data analytics to optimize the presentation. Therefore, a core problem with existing technologies is the inability to achieve real-time, personalized adjustments to video content to suit the interests and preferences of different visitors, thus impacting the overall quality of the visitor experience and the exhibition's appeal. Summary of the Invention

[0004] The purpose of this invention is to provide a food video display interactive system and method based on user feedback. By integrating visitor behavior records and sound signals to form comprehensive feedback data, it realizes dynamic updates of video content and loop optimization of playlists, solves the shortcomings of traditional food video display methods, significantly improves the interactivity and attractiveness of exhibitions, and provides visitors with a richer and more personalized visiting experience.

[0005] To achieve the above objectives, this invention proposes an interactive method for showcasing food videos based on user feedback, comprising the following steps:

[0006] Multiple screens are set up inside the exhibition hall, and food-related video materials are prepared in advance. Based on the entrance location selected by the visitor when entering the exhibition area, the corresponding preset video sequence is played on the screen.

[0007] The recording device records visitors' behavior and gaze direction, and adjusts the content of the video being played based on the recorded behavior.

[0008] Sound signals emitted by visitors are captured using sound sensing devices, and the sound signals are integrated with the behavioral records to form comprehensive feedback data.

[0009] The playlist is dynamically updated based on the feedback data, and the video display process is optimized in a loop to encourage visitors to stay longer and increase interaction.

[0010] Preferably, the installation of multiple screens within the exhibition hall and the pre-preparation of food-related video materials include:

[0011] The exhibition hall is divided into several exhibition areas, and a certain number of display screens are set up in each area;

[0012] For each display area, a set of images is arranged, which consists of a series of food-related images to ensure that the themes of the images are different between adjacent display areas;

[0013] Assign a single screen display duration to the image sequence, and adjust this value based on the average visitor dwell time;

[0014] Establish image switching rules to define how to smoothly transition from the current image to the next image when the duration of a single image display ends.

[0015] Preferably, the step of starting a corresponding preset image sequence to play on the screen based on the entrance location selected by the visitor when entering the exhibition area includes:

[0016] Determine the entrance location for visitors to the exhibition area and record this information as an initial reference point;

[0017] Based on the recorded entry location, select a matching set of images for preparation, ensuring that the sequence is associated with the display screen in the display area;

[0018] Set up a mapping relationship to match the selected video sequence with the entrance position one by one, so that each entrance has an independent video playlist;

[0019] When a visitor appears at the entrance, the corresponding video sequence is started according to the mapping relationship, so that the video content responds to the visitor's arrival in real time.

[0020] Preferably, the recording of visitors' lingering behavior and line of sight via the recording device includes:

[0021] The recording device is activated immediately after a visitor enters the exhibition area. The device tracks and captures the visitor's stopping position and eye movement path, forming raw behavioral trajectory data.

[0022] Based on the collected behavioral trajectory data, the duration of each visitor's stay in front of different images is analyzed, and this information is associated with the corresponding image sequence;

[0023] Based on the obtained dwell time and image association results, a basic dataset is prepared for subsequent image adjustments.

[0024] Preferably, adjusting the content of the playing video based on the behavior record includes:

[0025] The system receives and records information on visitor dwell time and gaze direction, which serves as the basis for adjusting the images.

[0026] Based on data on dwell time and gaze direction, determine the visitors' interests and select new video content that matches them;

[0027] Create update rules that guide the adjustment of image sequences based on identified interests, ensuring that new images match visitors' interests;

[0028] The update rules are applied to the currently playing video sequence to generate an adjusted video sequence, enabling real-time content updates.

[0029] Preferably, the method of capturing sound signals emitted by visitors using a sound sensing device includes:

[0030] Activate the sound sensing device to capture the sound signals emitted by visitors in the exhibition area, analyze the characteristics of the captured sound signals, identify sound patterns that express interest or participation, and mark them as sound feedback;

[0031] By combining audio feedback with the adjusted video sequence, the synergistic effect between the two is evaluated. Based on the obtained synergistic effect, a video fine-tuning scheme is developed to make subtle adjustments to the currently playing video sequence.

[0032] Preferably, the integration of the sound signal and the behavior record to form comprehensive feedback data includes:

[0033] Collect audio feedback, dwell time, and gaze direction information as the basic material for integration;

[0034] Construct a comprehensive feedback structure that pairs audio feedback with dwell time and gaze direction information to form an audiovisual association record of visitor behavior;

[0035] Based on the resulting comprehensive feedback structure, an engagement index is calculated for each exhibition area, which reflects the overall level of visitor interaction.

[0036] Based on the derived engagement metrics, a comprehensive feedback dataset was compiled, which includes engagement evaluation results for all display areas, and is used for subsequent image display optimization.

[0037] Preferably, the step of dynamically updating the playlist based on the generated feedback data includes:

[0038] Analyze the compiled comprehensive feedback dataset and extract the participation index for each display area;

[0039] Based on the extracted engagement metrics, assess the degree of match between the current playlist and visitors' interests, and identify content segments that need adjustment.

[0040] Develop an update strategy, based on the evaluation results, to determine how to select and prioritize new image content;

[0041] Apply update strategies to existing playlists to generate optimized playlists, ensuring that the new playlists dynamically reflect the latest visitor feedback.

[0042] Preferably, the step of performing a cyclical optimization of the image display process to encourage visitors to spend more time there and increase interaction includes:

[0043] Based on the generated optimized playlist, a new round of video display cycle will be started;

[0044] During the new round of exhibitions, we will continue to collect real-time feedback data from visitors, including new dwell times and audio feedback.

[0045] The collected real-time feedback data is compared and analyzed with the previous comprehensive feedback dataset to evaluate the changing trend of the display effect and record any significant changes;

[0046] Based on the observed trends, the update strategy was adjusted to form an improved update strategy, which guides the further optimization of subsequent playlists, ensuring that each loop responds to visitor interests and promotes increased dwell time and interaction.

[0047] On the other hand, the present invention proposes a user-feedback-based interactive system for showcasing food videos, comprising:

[0048] The exhibition area initialization and image sequence launch module is used to set up multiple screens in the exhibition hall and prepare food-related video materials in advance. Based on the entrance location selected by the visitor when entering the exhibition area, the corresponding preset image sequence is launched and played on the screen.

[0049] The behavior monitoring and image adjustment module is used to record the visitor's stopping behavior and gaze direction through a recording device, and adjust the content of the image being played based on the behavior record;

[0050] The sound feedback collection and integration module is used to capture the sound signals emitted by visitors using sound sensing devices, and integrate the sound signals with the behavior records to form comprehensive feedback data.

[0051] The playlist dynamic update and optimization module is used to dynamically update the playlist based on the generated feedback data, and to perform cyclical optimization of the video display process, encouraging visitors to stay longer and increase interaction.

[0052] Technical effects and advantages of the present invention: The food video display and interactive system and method based on user feedback proposed in this invention have the following advantages compared with the prior art:

[0053] This invention integrates visitor behavior records and audio signals to form comprehensive feedback data, enabling dynamic updates of video content and cyclical optimization of playlists. This method not only enhances the interactivity of the exhibition, allowing the content to respond to visitor interests and reactions in real time, but also effectively extends visitor dwell time, improving their visitor experience quality and satisfaction. By continuously optimizing the content, it ensures that each visit provides visitors with a fresh and personalized experience, significantly enhancing the overall attractiveness and interactivity of the exhibition. Attached Figure Description

[0054] Figure 1 This is a flowchart of the interactive food video display method based on user feedback according to the present invention;

[0055] Figure 2 This is a block diagram of the interactive food video display system based on user feedback according to the present invention. Detailed Implementation

[0056] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. The specific embodiments described herein are merely used to explain the present invention and are not intended to limit the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0057] This invention provides an interactive method for showcasing food videos based on user feedback. It aims to integrate visitor behavior records and audio signals to form comprehensive feedback data, thereby achieving dynamic updates of video content and cyclical optimization of playlists. This method can:

[0058] Enhanced interactivity: By capturing and analyzing visitors' behavior and reactions in real time, the system can quickly adjust the video content to make the display more relevant to visitors' interests and enhance the interactive experience.

[0059] Extend visitor dwell time: Personalized video displays are more likely to attract visitors' attention and increase the time they spend in the exhibition area.

[0060] Improve satisfaction: Continuously optimize the exhibition content based on actual visitor feedback to ensure that each visit brings visitors a fresh and interesting experience, thereby improving overall satisfaction.

[0061] In summary, this invention addresses the shortcomings of traditional food video display methods. By introducing a real-time feedback mechanism and intelligent optimization algorithms, it significantly enhances the interactivity and appeal of exhibitions, providing visitors with a richer and more personalized viewing experience. Specifically:

[0062] like Figure 1 As shown, the interactive food video display method based on user feedback in this invention includes the following steps:

[0063] Step 1: Set up multiple screens inside the exhibition hall and prepare food-related video materials in advance; specifically including:

[0064] The exhibition hall is divided into several display areas, with a certain number of display screens in each area. By rationally dividing the display areas and arranging the display screens, it is ensured that each area can independently display content on a specific theme, avoiding content repetition and enhancing the diversity of the visitor experience.

[0065] For each exhibition area, a video sequence Q is created, consisting of a series of food-related images to ensure that the themes of the images differ between adjacent exhibition areas. Different video sequences Q can provide diverse content to meet the needs of different visitors, and the thematic differentiation can guide visitors to move between different areas, increasing the choice and interest of the visit path.

[0066] Assign a time parameter T to the image sequence Q, where T represents the display duration of a single image. Adjust the value of T based on the average dwell time of visitors to ensure that each image has enough time to attract attention. The optimized time parameter T ensures that each image has enough display time to attract visitors' attention without being excessively prolonged, maintaining the rhythm of the display and improving the overall viewing experience.

[0067] An image transition rule R is established, which defines how to smoothly transition from the current image to the next image when T ends. Rule R depends on the set time parameter T to ensure a smooth and continuous presentation. The image transition rule R ensures a smooth transition during the presentation, avoiding the discomfort caused to visitors by sudden changes and enhancing visual continuity and comfort.

[0068] Formula expression: Let T = average dwell time / number of frames in image sequence Q. Let R(T) = if T expires, the next frame will be displayed. Q: image sequence, T: display time of a single frame, R: image switching rule.

[0069] T represents the duration of a single image display, calculated by dividing the average visitor's dwell time by the number of images in the image sequence Q. This ensures that each image has an appropriate display time, neither too fast nor too slow.

[0070] R defines the image switching logic, which means that when the display time T of a single screen expires, the next screen is automatically displayed. This rule ensures the automation and smoothness of the display process.

[0071] The following is the specific principle and implementation mechanism of image switching rule R:

[0072] 1. Dependence of the time parameter T

[0073] The image switching rule R strictly depends on the set time parameter T (the duration of a single image display). When the display time of the current image reaches the preset value T, the system will trigger the action of switching to the next image. This means that each image has a clear display cycle, ensuring the rhythm and consistency of the display process.

[0074] 2. Smooth transition mechanism

[0075] To avoid discomfort caused to visitors by sudden scene changes, the image transition rule R introduces a smooth transition mechanism. Specifically, this mechanism can include the following methods:

[0076] Fade-in / Fade-out: The current image gradually darkens and disappears, while the next image gradually brightens and appears. This method makes the transition between the two images appear natural and smooth.

[0077] Dissolve effect: The transition between the current frame and the next frame is achieved by pixel-level blending, producing a smooth transition effect.

[0078] Wipe effect: The new image gradually pushes away the old image from one side or corner until it completely covers the old image, giving a sense of advancement.

[0079] Animation effects such as rotation and scaling: Use 3D or 2D animation effects, such as rotation, scaling, and page turning, to make screen transitions more vivid and interesting.

[0080] 3. Automation Logic

[0081] The image switching rule R includes automation logic to ensure that the system can automatically switch images according to a preset time parameter T without human intervention. The key to this automation logic lies in precise timing and seamless transitions.

[0082] Precise timing: The system has an internal high-precision timer to track the actual display time of each frame. Once the time reaches the preset T value, the timer will immediately trigger a switching command.

[0083] Seamless transition: To ensure an uninterrupted switching process, the system prepares the content for the next screen just before the T-value expires and quickly executes the switching action when the T-value expires. This avoids any delays or stutters.

[0084] For example, suppose in a food exhibition hall, a sequence of images Q in one display area contains 10 images related to local specialty snacks. Preliminary research indicates that visitors spend an average of 5 minutes (300 seconds) in this display area. According to the formula above:

[0085] Let T = 300 seconds / 10 frames = 30 seconds. Let R (30 seconds) = if 30 seconds expires, then execute the next frame display;

[0086] This means each screen will be displayed for 30 seconds, after which the system will automatically switch to the next screen. This ensures that each screen has enough time to attract visitors' attention while maintaining the smoothness and rhythm of the entire presentation. Furthermore, if the average visitor dwell time in a particular area changes, for example, increasing to 6 minutes, the T-value can be recalculated by adjusting the formula to ensure optimal presentation results.

[0087] Step Two: Based on the entrance location selected by the visitor upon entering the exhibition area, the corresponding preset video sequence is played on the screen; specifically including:

[0088] The system identifies and records the entrance location for visitors to the exhibition area as an initial reference point. By recognizing and recording each visitor's entrance location, the system can create a personalized starting point for each visitor's experience. This not only improves the relevance and targeting of the exhibits but also enhances the interactivity and engagement of the visitor experience.

[0089] Based on the recorded entrance location, select a matching video sequence Q for preparation, ensuring that the sequence is linked to the display screens within the exhibition area; the video sequence Q corresponding to different entrance locations can provide the most relevant content based on the visitor's path preferences, ensuring that each exhibition area can independently display unique and coherent thematic content, avoiding repetition and monotony, and improving the overall quality of the visitor experience.

[0090] A mapping relationship M is set up to correspond the selected image sequence Q to the entrance location one by one, ensuring that each entrance has an independent image playlist; the image sequence Q corresponding to different entrance locations can provide the most relevant content according to the visitor's path preference, ensuring that each exhibition area can independently display unique and coherent thematic content, avoiding repetition and monotony, and improving the overall quality of the visitor experience.

[0091] When a visitor appears at the entrance, the corresponding video sequence Q is initiated based on the mapping relationship M, ensuring that the video content responds instantly to the visitor's arrival. This instant response mechanism ensures that the video content begins playing immediately upon the visitor's arrival, enhancing the immediacy and interactivity of the exhibition, leaving a lasting impression on the visitor, and increasing the overall appeal of the visit.

[0092] Formula expression: Let M = entrance position -> image sequence Q. Let P = current entrance position. If P corresponds to the position in M, then play M(P).

[0093] M represents the mapping relationship, defining the correspondence between each entrance location and a specific video sequence Q. P represents the current entrance location, i.e., the entrance the visitor actually enters. Q represents the video sequence, containing a series of food-related images. M(P) means finding the corresponding video sequence Q based on the current entrance location P by looking up the mapping relationship M, and then starting playback.

[0094] For example, suppose there are three different entrance locations A, B, and C in a food exhibition hall, each corresponding to a unique set of image sequences Q1, Q2, and Q3. The specific implementation is as follows:

[0095] Entrance Location A: When a visitor enters through entrance A, the system records this entrance location as an initial reference point.

[0096] Select image sequence Q1: Based on the recorded entry position A, the system selects the matching image sequence Q1 for preparation, ensuring that the sequence is associated with the display screen in the display area.

[0097] Set mapping relationship M: The system sets mapping relationship M, where A corresponds to Q1, B corresponds to Q2, and C corresponds to Q3.

[0098] Start the video sequence Q1: When a visitor appears at the entrance location A, the system starts playing the video sequence Q1 according to the mapping relationship M, so that the video content responds to the visitor's arrival in real time.

[0099] Let M = A->Q1, B->Q2, C->Q3, and let P = A (current entry position).

[0100] If P corresponds to position A in M, then play M(A), which is Q1.

[0101] This means that when visitors enter through entrance A, the system will automatically play video sequence Q1, providing unique content related to entrance A to enhance visitor engagement and experience quality. The same logic applies to other entrance locations B and C, playing corresponding video sequences Q2 and Q3 respectively.

[0102] Step 3: Record visitors' behavior and line of sight using recording devices; specifically including:

[0103] The recording devices are activated immediately upon visitors entering the exhibition area, ensuring no behavioral data is missed. This instant activation captures all behavioral data from the moment visitors enter the exhibition area, guaranteeing the completeness and accuracy of the records. This provides a comprehensive data foundation for subsequent analysis, improving the precision of adjustments to the exhibition content.

[0104] Recording devices are key tools for capturing and analyzing visitor behavior data. They consist of multiple components and technologies to ensure accurate and comprehensive recording of visitor behavior and gaze direction. The following is a detailed introduction to the specific components and working principles of the recording device:

[0105] 1. Hardware Components

[0106] Camera:

[0107] Function: Used to capture real-time video images within the exhibition area and track the movement paths of visitors.

[0108] Principle: High-definition cameras are installed at the top or side of the exhibition area, covering the entire exhibition space. Through video stream analysis technology, the movement trajectory of each visitor can be identified and tracked.

[0109] Infrared sensors:

[0110] Function: Detects the specific location and duration of a visitor's stay.

[0111] Principle: Infrared sensors are installed on the floor or walls of the exhibition area. When a visitor enters their sensing range, the sensor triggers a timer to start recording the dwell time. Simultaneously, through a multi-point sensor network, the visitor's location can be accurately determined.

[0112] Eye-tracking devices:

[0113] Function: Captures the direction of a visitor's gaze and their point of focus.

[0114] Principle: Non-invasive eye-tracking technology, such as infrared light emitters and high-resolution cameras, is used to capture eye movements by reflecting light, calculating the direction of gaze and the point of fixation. These devices are typically installed near the display screen to ensure accurate capture of changes in the visitor's gaze.

[0115] Radio Frequency Identification (RFID) Tags and Readers:

[0116] Function: To assist in identifying and tracking the behavior of specific visitors.

[0117] Principle: Visitors wear or carry cards or wristbands with RFID tags. When they enter the exhibition area, RFID readers installed in various locations automatically identify the tag information and record the visitor's identity and path. This helps to associate behavioral data with specific individuals.

[0118] Smart Sensor Nodes:

[0119] Function: Integrates multiple sensors to provide comprehensive data acquisition.

[0120] Principle: These nodes integrate multiple sensing devices such as cameras, infrared sensors, temperature and humidity sensors, and are connected to the central control system via a wireless network to achieve real-time data transmission and processing.

[0121] 2. Software and Technical Principles

[0122] Computer Vision:

[0123] Function: Extract and analyze visitor behavior patterns from video captured by cameras.

[0124] Principle: Utilizing advanced image processing algorithms, such as object detection, trajectory tracking, and pose estimation, the video stream is analyzed in real time to identify each visitor's actions, location, and movement path. Through deep learning models, it is also possible to predict visitors' points of interest and their next actions.

[0125] Data Analysis Platform:

[0126] Function: Integrates and processes data from various sensors to generate information such as behavioral trajectories and dwell time.

[0127] Principle: The data analysis platform receives raw data from cameras, infrared sensors, eye-tracking devices, etc., and constructs a detailed visitor behavior model through data cleaning, feature extraction, and pattern recognition. The platform also supports real-time data visualization, helping managers intuitively understand the exhibition effects.

[0128] Internet of Things (IoT) technology:

[0129] Function: Enables interconnection and interoperability between devices, ensuring real-time data transmission and synchronization.

[0130] Principle: All recording devices are connected to the central control system via wireless communication protocols such as Wi-Fi, Bluetooth, or Zigbee, forming a complete Internet of Things (IoT) network. This not only ensures the stability and efficiency of data transmission but also facilitates remote management and maintenance.

[0131] Suppose we are in a food exhibition hall, the recording equipment is used in the following specific ways:

[0132] Cameras and infrared sensors: A network of high-definition cameras and infrared sensors is installed at the top of the exhibition area. When a visitor enters an exhibition area, the system automatically starts recording, capturing their movement trajectory and dwell time.

[0133] Eye-tracking devices: Eye-tracking devices are installed next to each display screen to monitor visitors' gaze direction and fixation point in real time, and record their level of attention to different images.

[0134] RFID tags and readers: Each visitor is given a wristband with an RFID tag. When they enter the exhibition area, the RFID reader automatically identifies and records relevant information to ensure that behavioral data is linked to the individual.

[0135] Data Analysis Platform: All collected data is transmitted to the data analysis platform, where it is processed to generate detailed visitor behavior reports. For example, the system found that most visitors spent the longest time before image sequence Q2, indicating that this content was more popular, thus providing a basis for subsequent image adjustments.

[0136] By combining the aforementioned hardware and technologies, the recording equipment can not only comprehensively and accurately capture visitor behavior data, but also provide scientific data support for optimizing the exhibits, significantly improving the quality and interactivity of the visitor experience.

[0137] The system utilizes recording equipment to track and capture visitors' stopping positions and eye movement paths, forming raw behavioral trajectory data. By tracking these positions and eye movement paths, the system can obtain detailed visitor behavioral trajectory data. This data not only reflects visitors' specific behaviors in front of different images but also reveals their points of interest and focus, providing important evidence for personalized displays.

[0138] Based on the collected behavioral trajectory data, the system analyzes the duration (S) of each visitor's stay in front of different images and correlates this information with the corresponding image sequence (Q). The analysis of the stay duration (S) accurately measures the visitor's level of interest in each image. By correlating the stay duration (S) with the image sequence (Q), the system can identify which images are more attractive to visitors, thus providing a scientific basis for subsequent content optimization and ensuring that the displayed content better aligns with visitors' interests.

[0139] Based on the obtained dwell time S and image correlation results, a foundation dataset B is prepared for subsequent image adjustments to ensure that the adjusted images better align with visitor interests. Foundation dataset B integrates information from dwell time S and image sequence Q, providing solid data support for subsequent image adjustments. This dataset, based on actual behavior, makes image adjustments more targeted, significantly improving the relevance and attractiveness of the displayed content and enhancing the visitor's interactive experience.

[0140] Formula expression: Let S = dwell time, let Q = image sequence, let B(S,Q) = dataset, which contains a combination of dwell time S and corresponding image sequence Q.

[0141] S represents the dwell time, i.e., the time a visitor spends in front of a particular image. Q represents the image sequence, which contains a series of images related to food. B(S,Q) represents the dataset, consisting of the dwell time S and the corresponding image sequence Q, used for subsequent image adjustment analysis.

[0142] Dwell time (S) is a crucial indicator of a visitor's interest in a particular image. By recording behavioral trajectory data captured by the recording device, the system can accurately calculate the dwell time of each visitor in front of different images and associate these dwell times with specific image sequences (Q). The generated base dataset (B) not only records the viewing situation of each image but also reflects the visitor's preferences and points of interest, providing strong data support for subsequent image adjustments.

[0143] Suppose there are three different video sequences, Q1, Q2, and Q3, in a food exhibition hall, each showcasing different local specialty snacks. The implementation is as follows:

[0144] Activate recording equipment: When a visitor enters the exhibition area, the recording equipment is activated immediately to begin capturing their behavioral data.

[0145] Capturing behavioral trajectory data: The recording device tracks and captures the visitor's stopping positions and eye movement paths, forming raw behavioral trajectory data. For example, a visitor lingered for 45 seconds in front of image sequence Q1, 60 seconds in front of Q2, and 30 seconds in front of Q3.

[0146] Analysis of dwell time S: The system analyzes the collected behavioral trajectory data to calculate the dwell time S of each visitor in front of different images. For example:

[0147] Let S(Q1) = 45 seconds and S(Q2) = 60 seconds;

[0148] Let S(Q3) = 30 seconds.

[0149] Dataset B is created: Based on the correlation between dwell time S and image sequence Q, the system creates a basic dataset B for subsequent image adjustments. For example:

[0150] Let B(S,Q) = dataset B(45 seconds,Q1);

[0151] B(60 seconds, Q2);

[0152] B (30 seconds, Q3).

[0153] This means that if multiple visitors exhibit similar behavioral patterns (such as spending more time on Q2), the system can adjust the playlist based on this data, increasing the playback frequency of Q2 or extending its display time to better meet visitor interests. This video adjustment based on actual behavioral data can significantly improve the relevance and attractiveness of the displayed content, enhancing the visitor's interactive experience.

[0154] Step 4: Adjust the content of the playing video based on the recorded behavior; specifically including:

[0155] The system receives and records visitor dwell time (S) and gaze direction information, which serve as the basis for adjusting the visuals. By receiving detailed data on dwell time (S) and gaze direction, the system can comprehensively understand visitors' points of interest and behavioral patterns. This data provides a solid foundation for subsequent visual adjustments, ensuring that the adjusted visual content better meets the actual needs of visitors.

[0156] Based on visitor dwell time S and gaze direction data, the system determines visitors' interests and selects new video content Q' that matches them. Through in-depth analysis of dwell time S and gaze direction data, the system can accurately identify visitors' interests. Based on this understanding, selecting new video content Q' that matches their interests makes the displayed content more personalized and targeted, improving visitor engagement and satisfaction.

[0157] An update rule U is created, based on a defined interest preference, to guide the adjustment of the image sequence Q, ensuring that the new images align with the visitor's interests. Update rule U, formulated according to the visitor's interests, not only guides the direction of image sequence Q adjustments but also ensures the scientific rigor and effectiveness of the adjustment process. In this way, the system can dynamically respond to changes in visitor behavior, providing display content that better suits their preferences.

[0158] The update rule U is applied to the currently playing video sequence Q to generate an adjusted video sequence Q', enabling real-time content updates. By applying update rule U to the currently playing video sequence Q, the system can quickly generate the adjusted video sequence Q' and begin playback immediately. This not only ensures real-time updates of the displayed content but also enhances the interactivity and real-time nature of the presentation, providing visitors with a smoother and more personalized experience.

[0159] Formula expression: Let U(S,Q) = update rule, where S is the dwell time and Q is the original image sequence. Let Q' = U(S,Q), which represents the new image sequence after adjustment by update rule U.

[0160] S represents the dwell time, i.e., the time a visitor spends in front of a particular image. Q represents the original image sequence, which contains a series of food-related images. Q' represents the adjusted new image sequence, the image content after being adjusted by update rule U. U(S,Q) represents the update rule, which guides the adjustment of image content based on the dwell time S and the original image sequence Q.

[0161] The core of update rule U lies in dynamically adjusting the displayed content based on the visitor's dwell time S and the original video sequence Q, making it more aligned with the visitor's interests. Specifically, after receiving data on the visitor's dwell time S and gaze direction, the system analyzes this data to determine their interests and selects the most suitable new video content Q'. Then, the system applies this new content to the currently playing video sequence Q using update rule U, generating the adjusted video sequence Q', and immediately begins playback.

[0162] For example, suppose in a food exhibition hall, the image sequence Q in the display area includes three different image segments Q1, Q2, and Q3, each showcasing three different local specialty snacks. The specific implementation is as follows:

[0163] Received behavior records: The system received records of visitors staying in front of image sequence Q1 for 60 seconds, in front of Q2 for 45 seconds, and in front of Q3 for 30 seconds.

[0164] Identifying Interest Preferences: By analyzing the dwell time S and gaze direction data, the system found that visitors showed the highest interest (longest dwell time) in image sequence Q1. Therefore, the system determined that the visitor's interest leaned towards one of the local specialty snacks.

[0165] Create update rule U: Based on the above interest preferences, the system creates update rule U, which guides the adjustment of image sequence Q to add more image content related to local specialty snacks.

[0166] Apply update rule U: The system applies update rule U to the currently playing video sequence Q, generating an adjusted video sequence Q', for example:

[0167] Let U(60 seconds, Q1) = update rule;

[0168] Let Q' = U(60 seconds, Q1), representing the new image sequence adjusted by update rule U;

[0169] This means that if multiple visitors exhibit similar behavioral patterns (such as spending more time on Q1), the system can adjust the playlist based on this data, increasing the frequency of Q1-related content or extending its display time to better meet visitor interests. For example:

[0170] The original image sequence Q contains Q1, Q2 and Q3.

[0171] According to update rule U, the system added more content related to Q1 and generated a new image sequence Q', with the order becoming Q1, Q1 extended content, Q2, and Q3.

[0172] In this way, the system enables real-time updates of the displayed content, ensuring that each playback more accurately responds to visitors' interests, significantly improving the relevance and attractiveness of the display, and enhancing the interactive experience for visitors.

[0173] Step 5: Use sound sensing equipment to capture the sound signals emitted by visitors; specifically including:

[0174] Activating the sound sensors captures audio signals emitted by visitors within the exhibition area, ensuring coverage of all interactive zones. By activating the sound sensors, the system can capture various audio signals emitted by visitors within the exhibition area in real time. This not only expands the scope of interaction but also provides additional sources of behavioral data, enhancing the understanding of visitor interests and engagement.

[0175] Sound sensing devices are key tools for capturing and analyzing sound signals emitted by visitors within exhibition areas. Through various sensors and technologies, they accurately record and process diverse sound information, providing data support for subsequent behavioral analysis. The following is a detailed introduction to the specific structure and working principle of sound sensing devices:

[0176] 1. Hardware Components

[0177] Microphone Array:

[0178] Function: Used to capture sound signals within the display area.

[0179] Principle: Multiple high-sensitivity microphones are arranged in a specific geometric pattern at different locations within the display area, forming a microphone array. This layout enables sound source localization, accurately captures sound signals from different directions, and improves the signal-to-noise ratio.

[0180] Audio Amplifier:

[0181] Function: Enhances captured audio signals to ensure signal quality.

[0182] Principle: The audio amplifier amplifies the weak sound signal captured by the microphone array to a level suitable for subsequent processing. It also has a filtering function to reduce ambient noise interference.

[0183] Analog-to-Digital Converter (ADC):

[0184] Function: Converts captured analog audio signals into digital signals.

[0185] Principle: ADC quantizes analog audio signals into discrete digital values ​​at a certain sampling rate, making it easier for computer systems to process and analyze them.

[0186] Wireless Communication Module:

[0187] Function: Transmits the processed audio data to the central control system.

[0188] Principle: Using wireless communication protocols such as Wi-Fi, Bluetooth, or Zigbee, audio data is transmitted to the central control system in real time, ensuring the timeliness and reliability of the data.

[0189] Power Management Unit (PMU):

[0190] Function: Provides a stable power supply for the entire sound sensing device.

[0191] Principle: The PMU is responsible for managing the power input, ensuring that each component works in the best condition, and supporting low-power mode to extend the device's battery life.

[0192] 2. Software and Technical Principles

[0193] Sound Source Localization (SSL):

[0194] Function: Determines the exact location of the sound source.

[0195] Principle: By analyzing the time difference and intensity difference of the sound signals captured by the microphone array, algorithms such as Generalized Cross-Correlation Method (GCC-PHAT) or Beamforming are used to accurately locate the sound source.

[0196] Speech Recognition Technology:

[0197] Function: Identify and classify the various sounds made by visitors.

[0198] Principle: Using deep learning models (such as convolutional neural networks (CNNs) or recurrent neural networks (RNNs), feature extraction and pattern recognition are performed on the captured sound signals to distinguish different types of sounds (such as laughter, discussion, exclamations, etc.) and these sounds are labeled as sound feedback V.

[0199] Emotion Analysis Technology:

[0200] Function: Analyzes the emotional components in sound signals.

[0201] Principle: By analyzing the characteristics of voice such as tone, volume, and frequency changes, and combining them with machine learning models, the system identifies the emotional information contained in the voice (such as happiness, surprise, excitement, etc.), thereby enriching the content of the voice feedback.

[0202] Noise suppression and enhancement:

[0203] Function: Removes background noise and enhances useful sound signals.

[0204] Principle: By applying adaptive filters and spectral subtraction techniques, the impact of environmental noise is effectively reduced, the quality of the captured sound signal is improved, and the accuracy of subsequent analysis is ensured.

[0205] Real-time Data Analysis Platform:

[0206] Function: Integrates and processes data from multiple microphone arrays to generate comprehensive sound feedback reports.

[0207] Principle: The data analysis platform receives audio data from various microphone arrays and uses distributed computing and storage technologies to achieve real-time data processing and analysis. The platform also supports a visual interface to help managers intuitively understand the display effects.

[0208] Suppose a food exhibition hall is used, the specific application of sound sensing equipment is as follows:

[0209] Microphone array arrangement: Multiple microphone arrays are installed at the top and sides of the exhibition area, covering the entire exhibition space. Each array consists of eight high-sensitivity microphones to ensure the capture of sound signals from different directions.

[0210] Sound source localization: When a visitor stops in front of an image and makes a sound, the system uses sound source localization technology to accurately locate the sound source and record the specific image content corresponding to that location.

[0211] Sound Classification and Sentiment Analysis: The system classifies captured sound signals, identifying different types of sounds such as laughter, discussion, and exclamations. Through sentiment analysis technology, it determines the emotional components (such as happiness, surprise, etc.) expressed by these sounds. For example:

[0212] Let V = laughter, discussion, and exclamation; let emotion = happiness and surprise.

[0213] Real-time data transmission and processing: The captured sound signals are transmitted in real time to the central control system via a wireless communication module. They are then processed and analyzed on the data analysis platform to generate detailed sound feedback reports. For example:

[0214] Let Z = integrated feedback structure, which includes sound feedback V, dwell time S and gaze direction information Z(Q1') = (laughter, discussion, 60 seconds, focused gaze).

[0215] By combining the aforementioned hardware and technologies, sound sensing devices can not only comprehensively and accurately capture the sound signals emitted by visitors within the exhibition area, but also provide rich data support for subsequent behavioral analysis, significantly improving the effectiveness of exhibition content optimization and enhancing the interactive experience of visitors.

[0216] By analyzing the characteristics of the captured sound signals, sound patterns expressing interest or participation are identified and labeled as sound feedback V. Through characteristic analysis of the sound signals, the system can identify specific sound patterns (such as laughter, discussions, exclamations, etc.), which typically reflect the visitor's interest or participation. Labeling these patterns as sound feedback V provides important reference for subsequent image adjustments.

[0217] By combining the audio feedback V with the adjusted video sequence Q', the synergistic effect E between the two is evaluated to determine whether further optimization of the video content is needed. The system can assess the synergistic effect E by comparing the audio feedback V with the currently playing video sequence Q'. If a significant positive correlation is found between certain video content and the audio feedback V (i.e., visitors show greater interest when watching a particular video segment), it indicates that the video content is highly attractive, and vice versa. This evaluation helps determine whether further optimization of the video content is necessary.

[0218] Based on the derived synergy effect E, a video fine-tuning plan F is developed to make subtle adjustments to the currently played video sequence, enhancing the visitor's engagement experience. According to the evaluation results of synergy effect E, the system will develop video fine-tuning plan F to make subtle adjustments to the currently played video sequence. These adjustments aim to enhance the match between video content and visitor interests, thereby improving the overall engagement experience. For example, adding or extending popular content segments, and reducing or replacing less appealing parts.

[0219] Formula expression: Let V = sound feedback, let E(Q',V) = synergistic effect assessment, where Q' is the adjusted new image sequence, V is sound feedback, and let F(E) = image fine-tuning scheme, which is formulated based on the synergistic effect E.

[0220] V represents sound feedback, which is the sound pattern captured and analyzed by the sound sensing device. Q' represents the adjusted new image sequence, the image content after adjustment by update rule U. E(Q',V) represents the synergy assessment, used to measure the relationship between the adjusted image sequence Q' and the sound feedback V. F(E) represents the image fine-tuning scheme, formulated based on the synergy assessment result E, used to make minor adjustments to the image sequence.

[0221] Audio feedback (V) is a key indicator of visitor interest and engagement. By combining audio feedback (V) with the currently playing video sequence (Q'), the system can assess the synergistic effect (E) between the two. This assessment helps the system identify which video content is most likely to pique visitor interest and, based on this, develop video fine-tuning plans (F) to further optimize the presentation and enhance the visitor's experience.

[0222] For example, suppose in a food exhibition hall, the image sequence Q' in the display area includes three different image segments Q1', Q2', and Q3', each showing the preparation process of three different local specialty snacks. The specific implementation is as follows:

[0223] Activate sound sensors: Install multiple sound sensors in the exhibition area to ensure coverage of all interactive areas and capture sound signals emitted by visitors in real time.

[0224] Analyzing sound signal characteristics: The system analyzes the captured sound signals, identifies sound patterns expressing interest or participation (such as laughter, discussion, and exclamations), and labels these patterns as sound feedback V. For example:

[0225] Let V = laughter, discussion, and exclamation.

[0226] Assessing the synergistic effect E: The system combines the acoustic feedback V with the currently played video sequence Q' to assess the synergistic effect E between the two. For example, when playing video segment Q1', the system detected more laughter and discussion, indicating that the segment attracted high interest from visitors; while when playing Q3', there was less acoustic feedback, indicating lower appeal. Specific evaluation results are as follows:

[0227] Let E(Q1',V) = high synergy, E(Q2',V) = medium synergy, and E(Q3',V) = low synergy.

[0228] Develop image fine-tuning plan F: Based on the evaluation results of the synergy effect E, the system develops image fine-tuning plan F to make minor adjustments to the currently playing image sequence. For example, increase the playback frequency of Q1' or extend its display time, while reducing the playback time of Q3' or replacing its content. Specific adjustments are as follows:

[0229] Let F(E) = image fine-tuning scheme; F(high synergy) = increase the playback frequency of Q1' or extend the display time; F(medium synergy) = maintain the existing playback arrangement of Q2'; F(low synergy) = reduce the playback time of Q3' or replace its content.

[0230] This means that if the system observes similar behavioral patterns multiple times (such as Q1' evoking more laughter and discussion), it will dynamically adjust the playlist based on the results of the synergy effect E through the video fine-tuning scheme F, increasing popular content and reducing unpopular content, thereby better meeting the interests of visitors, significantly improving the relevance and attractiveness of the exhibition, and enhancing the interactive experience of visitors.

[0231] Step Six: Integrate the sound signals and the recorded behaviors to form comprehensive feedback data; specifically including:

[0232] The audio feedback (V), dwell time (S), and gaze direction information collected in step five are used as the foundational data for integration. By collecting audio feedback (V), dwell time (S), and gaze direction information, the system can acquire multi-dimensional visitor behavior data. This foundational data provides detailed information support for subsequent comprehensive analysis, ensuring the comprehensiveness and accuracy of the data, thereby improving the effectiveness of image display optimization.

[0233] A comprehensive feedback structure Z was constructed, pairing auditory feedback V with dwell time S and gaze direction information to form an audiovisual correlation record of visitor behavior. The comprehensive feedback structure Z integrates auditory feedback V, dwell time S, and gaze direction information, creating a multidimensional correlation record of visitor behavior. This structure not only reflects the specific behaviors of visitors in front of different images but also reveals their emotional responses and points of interest, laying the foundation for more in-depth behavioral analysis.

[0234] Based on the resulting comprehensive feedback structure Z, the participation index P for each exhibition area is calculated. This index reflects the overall level of visitor interaction. Through the comprehensive feedback structure Z, the system can calculate the participation index P for each exhibition area. This index quantifies the level of visitor interaction, including their dwell time, attention span, and emotional response in the area, providing a scientific basis for evaluating the effectiveness of the exhibition.

[0235] Based on the derived engagement index P, a comprehensive feedback dataset B was compiled. This dataset contains the engagement evaluation results for all display areas and is used for subsequent image display optimization. Comprehensive feedback dataset B integrates the engagement index P of each display area, forming a comprehensive dataset. This dataset not only records the engagement status of each area but also provides solid data support for subsequent image display optimization, making the optimization process more targeted and scientific.

[0236] Formula expression: Let Z = comprehensive feedback structure, which includes sound feedback V, dwell time S and gaze direction information. Let P(Z) = participation index, calculated based on the comprehensive feedback structure Z. Let B(P) = comprehensive feedback dataset, composed of the participation index P of each display area.

[0237] Z represents the comprehensive feedback structure, which includes audio feedback V, dwell time S, and gaze direction information, used to describe visitor behavior and reactions. P(Z) represents the engagement index, calculated based on the comprehensive feedback structure Z, reflecting the overall interaction level of visitors in a specific exhibition area. B(P) represents the comprehensive feedback dataset, composed of the engagement index P of each exhibition area, used for subsequent image display optimization.

[0238] The comprehensive feedback structure Z integrates audio feedback (V), dwell time (S), and gaze direction information to form a multi-dimensional behavioral record. Through comprehensive analysis of this data, the system can calculate the engagement index (P) for each display area, thus gaining a comprehensive understanding of visitor interaction. Finally, based on these engagement indices P, the system compiles a comprehensive feedback dataset B, providing data support for optimizing video presentations.

[0239] Suppose that in a food exhibition hall, there are three different video sequences Q1', Q2', and Q3' in the display area, each showcasing the preparation process of three local specialty snacks. The specific implementation is as follows:

[0240] Collect basic materials: The system collects sound feedback V (such as laughter, discussion, and exclamations) from sound sensing devices, as well as the duration S of visitors' stay in front of different images and their gaze direction.

[0241] Let V = laughter, discussion, and exclamations;

[0242] Let S(Q1') = 60 seconds, S(Q2') = 45 seconds, and S(Q3') = 30 seconds;

[0243] Constructing a comprehensive feedback structure Z: The system pairs sound feedback V, dwell time S, and gaze direction information to form a comprehensive feedback structure Z. For example, when playing Q1', more laughter and discussion are detected, and the dwell time is longer; while when playing Q3', there is less sound feedback and the dwell time is shorter.

[0244] Let Z = (V, S, line-of-sight information);

[0245] Z(Q1') = (laughter, discussion, 60 seconds, focused gaze);

[0246] Z(Q2') = (slight discussion, 45 seconds, moderate gaze);

[0247] Z(Q3') = (few exclamations, 30 seconds, distracted gaze);

[0248] Calculate the engagement metric P: Based on the comprehensive feedback structure Z, the system calculates the engagement metric P for each display area. For example:

[0249] Let P(Z) = participation index;

[0250] P(Z(Q1')) = High participation;

[0251] P(Z(Q2')) = Moderate participation;

[0252] P(Z(Q3')) = Low participation;

[0253] Compile a comprehensive feedback dataset B: Based on the obtained participation index P, the system compiles a comprehensive feedback dataset B, which contains the participation evaluation results of all display areas and is used for subsequent image display optimization.

[0254] Let B(P) = comprehensive feedback dataset;

[0255] B (High Engagement, Q1');

[0256] B (Medium participation, Q2');

[0257] B (Low participation, Q3');

[0258] This means that if multiple display areas exhibit similar behavioral patterns (such as Q1' having high engagement), the system can adjust the playlist based on this data, increasing popular content and reducing unpopular content, thereby better meeting visitors' interests, significantly improving the relevance and attractiveness of the display, and enhancing visitors' interactive experience.

[0259] Step 7: Dynamically update the playlist based on the generated feedback data; specifically including:

[0260] The system analyzes and compiles a comprehensive feedback dataset B, extracting the participation index P for each exhibition area. Through detailed analysis of the comprehensive feedback dataset B, the system can extract the participation index P for each exhibition area. These indicators quantify the interaction level of visitors in different areas, providing a scientific basis for subsequent evaluation and adjustments, ensuring that the optimization process is targeted and effective.

[0261] Based on the extracted engagement metric P, the system assesses the match between the current playlist L and visitor interests, identifying content segments requiring adjustment. By comparing engagement metric P with the current playlist L, the system evaluates the degree of match between the two. If certain content segments are found to have low engagement, these segments are marked as requiring adjustment. This assessment helps identify shortcomings in the presented content, allowing for targeted optimization.

[0262] An update strategy G is developed, which, based on the evaluation results, determines how to select and prioritize new video content Q' to better respond to visitor preferences. Update strategy G, developed based on the evaluation results, not only guides the selection of new video content Q' but also determines its order within the playlist. In this way, the system can more accurately meet visitor interests and preferences, enhancing the relevance and appeal of the displayed content.

[0263] The update strategy G is applied to the existing playlist L to generate an optimized playlist L', ensuring that the new playlist dynamically reflects the latest visitor feedback. By applying update strategy G to the existing playlist L, the system can quickly generate the optimized playlist L'. This not only guarantees the real-time updating of the displayed content but also enhances the interactivity and real-time nature of the presentation, making each playback more aligned with the latest visitor feedback and providing a more personalized experience.

[0264] Formula expression: Let G(P) = update strategy. Based on the participation index P, let L' = G(L), which represents the new playlist after optimizing the original playlist L through the update strategy G.

[0265] G(P) represents the update strategy, formulated based on the engagement metric P, which guides the selection and sorting of new video content Q'. L represents the original playlist, containing all currently displayed video sequences. L' represents the optimized playlist, the latest playlist adjusted by the update strategy G.

[0266] The core of the update strategy G lies in assessing the match between the current playlist L and visitor interests based on the engagement metric P, and selecting and prioritizing new video content Q' accordingly. Specifically, the system analyzes the engagement metric P for each display area, identifies low-engagement content segments, and replaces or adjusts these segments using the update strategy G, generating an optimized playlist L'. This dynamic update mechanism ensures that the displayed content remains aligned with visitor interests, enhancing the overall interactive experience.

[0267] For example, suppose in a food exhibition hall, there are three different video sequences Q1, Q2, and Q3 in the display area, each showing the preparation process of three local specialty snacks. The specific implementation is as follows:

[0268] Analyzing the Comprehensive Feedback Dataset B: The system analyzes the compiled comprehensive feedback dataset B and extracts the participation index P for each display area. For example:

[0269] Let P(Q1) = high participation;

[0270] Let P(Q2) = moderate participation;

[0271] Let P(Q3) = low participation;

[0272] Evaluating Playlist L: The system assesses the match between the current playlist L and the visitors' interests based on the extracted engagement metric P. For example, it finds that Q3 has low engagement and needs adjustment. Specific evaluation results are as follows:

[0273] Let L = [Q1, Q2, Q3];

[0274] The section marked Q3 is the content that needs to be adjusted;

[0275] Develop an update strategy G: Based on the evaluation results, the system develops an update strategy G, selecting and prioritizing new video content Q' to better respond to visitor preferences. For example, it adds more videos related to high-engagement content (such as Q1) and reduces or replaces low-engagement content (such as Q3). Specific strategies are as follows:

[0276] Let G(P) = update policy;

[0277] G (High Engagement) = Increase the amount of video footage related to Q1;

[0278] G (Low Participation) = Reduce or replace Q3;

[0279] Generate an optimized playlist L': The system applies the update strategy G to the existing playlist L, generating an optimized playlist L' that dynamically reflects the latest visitor feedback. For example:

[0280] Let L' = G(L);

[0281] L'=[Q1,Q1 extended content,Q2,new Q3 content];

[0282] This means that if multiple display areas exhibit similar behavioral patterns (such as Q1 having high engagement), the system can adjust the playlist based on this data, increasing popular content and reducing unpopular content, thereby better meeting visitors' interests, significantly improving the relevance and attractiveness of the display, and enhancing visitors' interactive experience.

[0283] Step 8: Optimize the video presentation flow in a loop to encourage visitors to stay longer and increase interaction; specifically including:

[0284] Based on the generated optimized playlist L', a new round of video presentation cycle is initiated. By applying the optimized playlist L', the system can initiate a new round of video presentation cycle. This not only ensures the real-time updating of the presentation content but also provides a new foundation for subsequent feedback collection and analysis, enabling each presentation to more accurately respond to visitors' interests.

[0285] During the new round of exhibitions, real-time feedback data R from visitors is continuously collected. This data includes the new dwell time S' and audio feedback V'. By continuously collecting real-time feedback data R (including the new dwell time S' and audio feedback V'), the system can dynamically monitor changes in visitor behavior and obtain the latest interaction information. This real-time data provides detailed information support for subsequent evaluation and optimization, ensuring that the adjustment process is targeted and effective.

[0286] The system compares and analyzes the collected real-time feedback data R with the previous comprehensive feedback dataset B to evaluate the trend T of the display effect and records any significant changes. This comparison not only reveals the effectiveness of the content improvement but also identifies areas requiring further optimization, providing a basis for developing more effective update strategies.

[0287] Based on the derived trend T, the update strategy G is adjusted to form an improved update strategy G', which guides further optimization of subsequent playlists, ensuring that each loop more accurately responds to visitor interests and promotes increased dwell time and interaction. According to the changing trend T, the system adjusts the original update strategy G to form an improved update strategy G'. This not only improves the match between the displayed content and visitor interests but also enhances the interactivity and appeal of the display, ensuring that each loop better meets visitor expectations, thereby extending dwell time and increasing interaction.

[0288] Formula expression: Let R = real-time feedback data, including the new dwell time S' and sound feedback V'. Let T(R,B) = trend of change, and compare the real-time feedback data R with the comprehensive feedback dataset B to obtain the improved update strategy, which is adjusted according to the trend of change T.

[0289] L': Optimized playlist, R: Real-time feedback data, S': New dwell time, V': New sound feedback, B: Comprehensive feedback dataset, T: Change trend, G: Original update strategy, G': Improved update strategy.

[0290] Real-time feedback data R is a crucial indicator for measuring the effectiveness of a presentation. By comparing and analyzing the real-time feedback data R with the previous comprehensive feedback dataset B, the system can assess the changing trends T of the presentation's effectiveness. Based on these trends, the system adjusts the original update strategy G, resulting in an improved update strategy G', to achieve continuous optimization of the presentation content. This method ensures that each iteration responds more accurately to visitors' interests, thereby extending dwell time and increasing interaction.

[0291] For example, suppose in a food exhibition hall, there are three different video sequences Q1, Q2, and Q3 in the display area, each showing the preparation process of three local specialty snacks. The specific implementation is as follows:

[0292] Start a new video display cycle: The system starts a new video display cycle based on the generated optimized playlist L'. For example, the new playlist L' is [Q1, Q1 extended content, Q2, new Q3 content].

[0293] Let L' = [Q1, Q1 extended content, Q2, new Q3 content].

[0294] Continuously collect real-time feedback data R: During a new round of exhibitions, the system continuously collects real-time feedback data R from visitors, including new dwell time S' and audio feedback V'. For example:

[0295] Let R = (S', V') S'(Q1) = 70 seconds, S'(Q2) = 50 seconds, S'(new Q3 content) = 40 seconds, V'(Q1) = more laughter and discussion, V'(Q2) = moderate discussion, V'(new Q3 content) = a few exclamations.

[0296] Comparative analysis to assess the trend T: The system compares the collected real-time feedback data R with the previous comprehensive feedback dataset B to assess the trend T of the display effect. For example, it was found that the participation rate for Q1 further increased, while the participation rate for the new Q3 content remained low. Specific evaluation results are as follows:

[0297] Let T(R,B) = trend of change, and T(Q1) = significant increase in participation.

[0298] T (New Q3 Content) = No significant improvement in engagement.

[0299] Adjust the update strategy G to form an improved update strategy G': Based on the derived trend T, the system adjusts the original update strategy G to form an improved update strategy G', to guide further optimization of the playlist. For example, add more relevant footage about Q1 and continue to improve the new Q3 content. The specific strategies are as follows:

[0300] Let G'(T) = Improved update strategy; G'(high engagement increase) = Add more relevant images about Q1; G'(low engagement no improvement) = Further improve the new Q3 content or consider replacement.

[0301] This means that if multiple display areas exhibit similar behavioral patterns (such as Q1 having higher engagement), the system can adjust the playlist based on this data, increasing popular content and reducing unpopular content, thereby better meeting visitors' interests, significantly improving the relevance and attractiveness of the display, and enhancing visitors' interactive experience.

[0302] In this way, the system continuously optimizes the displayed content, ensuring that each playback more accurately responds to visitors' interests, providing a more personalized and engaging display experience, and ultimately promoting increased dwell time and interaction.

[0303] On the other hand, this invention proposes a user-feedback-based interactive system for showcasing food videos, such as... Figure 2 Shown, including:

[0304] The exhibition area initialization and image sequence launch module is used to set up multiple screens in the exhibition hall and prepare food-related video materials in advance. Based on the entrance location selected by the visitor when entering the exhibition area, the corresponding preset image sequence is launched and played on the screen.

[0305] The behavior monitoring and image adjustment module is used to record the visitor's stopping behavior and gaze direction through a recording device, and adjust the content of the image being played based on the behavior record;

[0306] The sound feedback collection and integration module is used to capture the sound signals emitted by visitors using sound sensing devices, and integrate the sound signals with the behavior records to form comprehensive feedback data.

[0307] The playlist dynamic update and optimization module is used to dynamically update the playlist based on the generated feedback data, and to perform cyclical optimization of the video display process, encouraging visitors to stay longer and increase interaction.

[0308] In addition, the aforementioned display area initialization and image sequence startup module, behavior monitoring and image adjustment module, sound feedback collection and integration module, and playlist dynamic update and optimization module are also used to implement other steps of the aforementioned user feedback food video display interaction method, which will not be elaborated here.

[0309] In summary, this invention integrates visitor behavior records and audio signals to form comprehensive feedback data, enabling dynamic updates of video content and cyclical optimization of playlists. This method not only enhances the interactivity of the exhibition, allowing the displayed content to respond to visitors' interests and reactions in real time, but also effectively extends visitors' dwell time, improving their visitor experience quality and satisfaction. By continuously optimizing the displayed content, ensuring that each visit brings visitors a fresh and personalized experience, the overall attractiveness and interactivity of the exhibition are significantly enhanced.

[0310] Finally, it should be noted that the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for interactive food video presentation based on user feedback, characterized in that, Includes the following steps: Multiple screens are set up within the exhibition hall, and food-related video materials are prepared in advance. Based on the entrance location selected by visitors when entering the exhibition area, a corresponding preset video sequence is played on the screen. This includes: determining the visitor's entrance location and recording this information as an initial reference point; selecting and preparing a matching video sequence based on the recorded entrance location, ensuring that the sequence is linked to the display screen within the exhibition area; setting a mapping relationship to correspond the selected video sequence one-to-one with the entrance location, ensuring that each entrance has an independent video playlist; when a visitor appears at the entrance location, the corresponding video sequence is started according to the mapping relationship, so that the video content responds instantly to the visitor's arrival. The process involves recording visitor behavior and gaze direction using a recording device. This includes: activating the recording device immediately upon a visitor entering the exhibition area; tracking and capturing the visitor's stopping position and gaze path to form raw behavioral trajectory data; analyzing the duration of each visitor's stay in front of different images based on the collected behavioral trajectory data, and associating this information with the corresponding image sequence; preparing a basic dataset for subsequent image adjustments based on the obtained stay duration and image association results; and adjusting the content of the currently playing image based on the behavioral records. Sound signals emitted by visitors are captured using sound sensing devices, and the sound signals are integrated with the behavioral records to form comprehensive feedback data. The playlist is dynamically updated based on the feedback data, and the video display process is optimized in a loop to encourage visitors to stay longer and increase interaction. The plan involves setting up multiple screens within the exhibition hall and preparing food-related video materials in advance, including: The exhibition hall is divided into several exhibition areas, and a certain number of display screens are set up in each area; For each display area, a set of images is arranged, which consists of a series of food-related images to ensure that the themes of the images are different between adjacent display areas; Assign a single screen display duration to the image sequence, and adjust this value based on the average visitor dwell time; Establish image transition rules to define how to smoothly transition from the current image to the next image when the duration of a single image display ends; Adjusting the content of the playing video based on the behavior record includes: The system receives and records information on visitor dwell time and gaze direction, which serves as the basis for adjusting the images. Based on data on dwell time and gaze direction, determine the visitors' interests and select new video content that matches them; Create update rules that guide the adjustment of image sequences based on identified interests, ensuring that new images match visitors' interests; The update rules are applied to the currently playing video sequence to generate an adjusted video sequence, enabling real-time content updates. The method of capturing sound signals emitted by visitors using sound sensing devices includes: Activate the sound sensing device to capture the sound signals emitted by visitors in the exhibition area, analyze the characteristics of the captured sound signals, identify sound patterns that express interest or participation, and mark them as sound feedback; By combining audio feedback with the adjusted video sequence, the synergistic effect between the two is evaluated. Based on the obtained synergistic effect, a video fine-tuning scheme is formulated to make subtle adjustments to the currently playing video sequence. The integration of the sound signal and the behavior record to form comprehensive feedback data includes: Collect audio feedback, dwell time, and gaze direction information as the basic material for integration; Construct a comprehensive feedback structure that pairs audio feedback with dwell time and gaze direction information to form an audiovisual association record of visitor behavior; Based on the resulting comprehensive feedback structure, an engagement index is calculated for each exhibition area, which reflects the overall level of visitor interaction. Based on the derived engagement metrics, a comprehensive feedback dataset was compiled, which includes engagement evaluation results for all display areas, and is used for subsequent image display optimization.

2. The method for interactive food video display based on user feedback according to claim 1, characterized in that, The dynamic updating of the playlist based on the generated feedback data includes: Analyze the compiled comprehensive feedback dataset and extract the participation index for each display area; Based on the extracted engagement metrics, assess the degree of match between the current playlist and visitors' interests, and identify content segments that need adjustment. Develop an update strategy, based on the evaluation results, to determine how to select and prioritize new image content; Apply update strategies to existing playlists to generate optimized playlists, ensuring that the new playlists dynamically reflect the latest visitor feedback.

3. The method for interactive food video display based on user feedback according to claim 2, characterized in that, The aforementioned process of cyclically optimizing the image display workflow to encourage visitors to spend more time there and increase interaction includes: Based on the generated optimized playlist, a new round of video display cycle will be started; During the new round of exhibitions, we will continue to collect real-time feedback data from visitors, including new dwell times and audio feedback. The collected real-time feedback data is compared and analyzed with the previous comprehensive feedback dataset to evaluate the changing trend of the display effect and record any significant changes; Based on the observed trends, the update strategy was adjusted to form an improved update strategy, which guides the further optimization of subsequent playlists, ensuring that each loop responds to visitor interests and promotes increased dwell time and interaction.

4. A food video display and interactive system for performing user feedback as described in any one of claims 1-3, characterized in that, include: The exhibition area initialization and image sequence activation module is used to set up multiple screens within the exhibition hall and pre-prepare food-related video materials. Based on the entrance location selected by visitors upon entering the exhibition area, it activates a corresponding preset image sequence to play on the screen. This includes: determining the visitor's entrance location and recording this information as an initial reference point; selecting and preparing a matching image sequence based on the recorded entrance location, ensuring the sequence is linked to the display screen within the exhibition area; setting a mapping relationship to correspond the selected image sequence one-to-one with the entrance location, ensuring each entrance has an independent image playlist; and activating the corresponding image sequence based on the mapping relationship when a visitor appears at the entrance location, ensuring the video content responds instantly to the visitor's arrival. The behavior monitoring and image adjustment module is used to record visitors' lingering behavior and gaze direction through a recording device. This includes: activating the recording device, which is activated immediately upon a visitor entering the exhibition area; tracking and capturing the visitor's lingering position and gaze movement path to form raw behavior trajectory data; analyzing the duration of each visitor's stay in front of different images based on the collected behavior trajectory data, and associating this information with the corresponding image sequence; preparing a basic dataset for subsequent image adjustment based on the obtained lingering duration and image association results; and adjusting the content of the currently playing image based on the behavior records. The sound feedback collection and integration module is used to capture the sound signals emitted by visitors using sound sensing devices, and integrate the sound signals with the behavior records to form comprehensive feedback data. The playlist dynamic update and optimization module is used to dynamically update the playlist based on the generated feedback data, and to perform cyclical optimization of the video display process, encouraging visitors to stay longer and increase interaction.

Images