Rapid drawing and word guessing system based on artificial intelligence

Abstract

The invention relates to the technical field of artificial intelligence drawing, in particular to a rapid drawing and word guessing system based on artificial intelligence, and the system obtains user input and game data through an acquisition module, a processing module generates an image through the artificial intelligence and calculates the semantic matching degree, and a control module compares the matching degree and dynamically adjusts the game difficulty. And the feedback module provides personalized feedback and updates the word bank. The system realizes game personalization, intelligent matching and adaptive difficulty adjustment, and improves user experience and game interactivity.

Description

Technical Field

[0001] The present invention relates to the field of artificial intelligence painting technology, and in particular to an artificial intelligence-based fast painting word guessing system. Background Art

[0002] Before the development of artificial intelligence technology, picture-guessing games were mainly played offline, such as "You Draw, I Guess." These games usually required players to have certain drawing skills and language expression abilities to effectively communicate and guess words.

[0003] With the development of artificial intelligence, image processing technology continues to advance, including image enhancement, filtering, and feature extraction, providing better input materials and processing methods for AI painting. At the same time, the rise of computer vision enables computers to understand and interpret image content, providing AI painting with a deeper level of visual understanding.

[0004] Patent document with publication number CN116342739A discloses a method, electronic device and medium for generating multiple painting images based on artificial intelligence. The method includes: when any display device among multiple display devices receives a voice painting instruction, extracting painting elements from the voice painting instruction; based on a first artificial intelligence model, generating a first painting image using the painting elements; based on a second artificial intelligence model, generating a second painting image using the first painting image; displaying the first painting image on a master device among the multiple display devices, and displaying the second painting image on a slave device among the multiple display devices.

[0005] This shows that the following problem exists: the prior art does not provide a mechanism for dynamically adjusting game parameters, resulting in a lack of challenge and fun. Summary of the Invention

[0006] To this end, the present invention provides a fast drawing and word guessing system based on artificial intelligence to overcome the problem in the prior art that the game lacks challenge and fun due to the lack of a mechanism for dynamically adjusting game parameters.

[0007] To achieve the above object, the present invention provides a fast picture-guessing system based on artificial intelligence, comprising:

[0008] The acquisition module is used to obtain the initial input information of each user in real time, including the initial word library type, game mode and difficulty level, and is also used to obtain the guessing input language and actual guessing time of each user in real time;

[0009] a processing module, connected to the acquisition module, configured to generate an initial image based on artificial intelligence technology and according to the initial vocabulary type, the difficulty level, and the game mode, and to analyze the semantic relevance between the initial image and the guessing input language according to a predefined model to obtain an actual similarity match;

[0010] a control module, connected to the processing module, configured to compare the actual similarity matching degree with a preset standard similarity matching degree to obtain an initial matching result, determine a target matching result based on the initial matching result, the actual guessing time, and the preset standard guessing time, and adjust the difficulty level based on the target matching result and the game mode;

[0011] A feedback module is connected to the control module, and is used to feed back the target matching result to the user, and provide different feedback information according to the game mode, record the feedback information to obtain a record result, analyze the record result to update the initial vocabulary type, and then obtain the target vocabulary type.

[0012] Furthermore, the processing module includes:

[0013] an image generating unit, which generates an image matching the initial vocabulary type and the difficulty level using an artificial intelligence algorithm according to the initial vocabulary type, the game mode, and the difficulty level to obtain the initial image;

[0014] a feature extraction unit, connected to the image generation unit, for extracting color, shape and texture from the initial image to obtain a first feature result;

[0015] a semantic analysis unit connected to the feature extraction unit, configured to convert the guessing word input into a semantic representation that matches the feature result to obtain a semantic result;

[0016] A similarity calculation unit is connected to the semantic analysis unit and is used to analyze the correlation between the feature result and the semantic result according to the predefined model to obtain the actual similarity matching degree.

[0017] Furthermore, the semantic unit includes:

[0018] A text preprocessing subunit, for cleaning and standardizing the guessing word input to obtain a processing result;

[0019] an embedding generation subunit, connected to the text preprocessing subunit, for converting the processing result into a vector representation to obtain a conversion result;

[0020] a feature extraction subunit, connected to the embedding generation subunit, for inputting the conversion result into a deep learning model to extract semantic features to obtain a second feature result;

[0021] The semantic generation subunit is connected to the feature extraction subunit and is used to aggregate the second feature result into a vector of a fixed length to obtain the semantic result.

[0022] Furthermore, the similarity calculation unit includes:

[0023] A vector comparison subunit, configured to compare the feature result and the semantic result to calculate the similarity between them and obtain a calculation result;

[0024] a similarity scoring subunit, connected to the vector comparison subunit, for assigning a similarity score to each pair of image features and text semantics according to the calculation results to obtain a plurality of score results;

[0025] The result aggregation subunit is connected to the similarity scoring subunit and is used to aggregate or average a plurality of the score results to obtain the actual similarity matching degree.

[0026] Furthermore, the control module includes:

[0027] a matching evaluation unit, configured to compare the actual similarity matching degree with the standard similarity matching degree to obtain the initial matching result;

[0028] a time evaluation unit, configured to compare the actual guessing time with the standard guessing time to evaluate the user's guessing efficiency and thereby obtain a time evaluation result;

[0029] a result determination unit, connected to the matching evaluation unit and the time evaluation unit respectively, for determining the target matching result according to the initial matching result and the time evaluation result;

[0030] A difficulty adjustment unit is connected to the result determination unit and is used to adjust the difficulty level according to the target matching result and the game mode.

[0031] Furthermore, the time evaluation unit includes:

[0032] A time acquisition subunit, for acquiring or setting a standard guessing time for each word in the game to obtain the standard guessing time;

[0033] a time comparison subunit, connected to the time acquisition subunit, for comparing the actual word guessing time with the standard word guessing time to obtain a comparison result;

[0034] The efficiency evaluation subunit is connected to the time comparison subunit and is used to evaluate the word guessing efficiency according to the comparison result to obtain the time evaluation result.

[0035] Furthermore, the difficulty adjustment unit includes:

[0036] a difficulty analysis subunit, configured to analyze the target matching result and the impact of the game mode on the difficulty level to obtain an impact result;

[0037] The difficulty adjustment subunit is connected to the difficulty analysis subunit and is used to adjust the difficulty level according to the impact result.

[0038] Furthermore, the feedback module includes:

[0039] An information generating unit, configured to generate corresponding feedback information according to the game difficulty and the target matching result to obtain a feedback result;

[0040] A feedback recording unit, connected to the information generating unit, for storing the feedback result in a database or a log file to obtain a record result;

[0041] an information analysis unit, connected to the feedback recording unit, for analyzing the user behavior pattern and difficulty adaptability according to the recording result to obtain the analysis result;

[0042] The word library updating unit is connected to the data analyzing unit and is used to update the initial word library type according to the analysis result.

[0043] Furthermore, the information analysis unit includes:

[0044] a behavior analysis subunit, configured to analyze a user's behavior pattern to obtain a first analysis result;

[0045] a difficulty analysis subunit, configured to analyze the user's adaptability to the difficulty level to obtain a second analysis result;

[0046] The result synthesis subunit is connected to the behavior analysis subunit and the difficulty analysis subunit respectively, and is used to synthesize the first analysis result and the second analysis result to obtain the analysis result.

[0047] Furthermore, the vocabulary updating unit includes:

[0048] a vocabulary screening subunit, configured to screen out updated vocabulary based on the analysis result to obtain a screening result;

[0049] a difficulty adjustment subunit, connected to the vocabulary screening subunit, for adjusting the difficulty level of the screening result to obtain an adjusted result;

[0050] The vocabulary verification subunit is connected to the difficulty adjustment subunit and is used to verify whether the adjustment result is suitable for the user group.

[0051] Compared with the prior art, the beneficial effect of the present invention is that the present invention obtains the user's guessing input language and guessing time in real time through the acquisition module, thereby enhancing the interactivity and immediacy of the game and making the user experience smoother and more interesting. The processing module uses artificial intelligence technology to generate images, ensuring the diversity and creativity of the images, and improving the playability and attractiveness of the game. By analyzing the semantic relevance between the image and the guessing input language through a predefined model, a more accurate similarity match can be provided to ensure the fairness of the game results. The control module dynamically adjusts the difficulty level according to the user's actual performance (similarity match and guessing time) to ensure that the game is appropriately challenging for the user. The feedback module increases the interactivity of the game and the user's sense of participation by providing real-time feedback on the target matching results to the user.

[0052] In particular, the image generation unit can create customized image content that matches the game theme and difficulty level by generating images based on the initial vocabulary type, game mode, and difficulty level. The feature extraction unit converts image data into feature vectors, simplifying the data structure and facilitating computer processing and analysis. The semantic analysis unit can more accurately match user input with image content through semantic analysis, improving the game's responsiveness. The similarity calculation unit can quantify the degree of match between the user's guess and the image by analyzing the correlation between feature results and semantic results using a predefined model.

[0053] In particular, the text preprocessing subunit reduces noise by removing irrelevant characters and stop words, improving the accuracy of subsequent processing steps. Converting text to a unified format ensures that the embedding generation subunit can consistently process input data. The embedding generation subunit converts text into vectors, enabling computers to better understand and process natural language information. Through model learning, the feature extraction subunit abstracts raw text into higher-level feature representations for easier understanding and analysis. The semantic generation subunit compresses and represents raw text information through aggregation operations, which helps improve computational efficiency and reduce storage costs.

[0054] In particular, the vector comparison subunit calculates the similarity between image feature vectors and text semantic vectors, quantifying their proximity in vector space. The similarity calculation subunit clearly visualizes the degree of similarity between different images and texts through scores, helping to distinguish between highly and less similar pairs.

[0055] In particular, the match evaluation unit accurately assesses the accuracy of a user's guess by comparing the actual similarity match with the standard similarity match. The time evaluation unit measures the user's reaction speed and efficiency by comparing the actual guess time with the standard guess time. The result determination unit dynamically adjusts the game based on the user's overall performance to maintain its challenging and interesting gameplay. The difficulty adjustment unit dynamically adjusts the difficulty based on the user's gaming performance and gaming patterns to maintain the game's appeal.

[0056] In particular, the time acquisition subunit sets a standard guessing time for each word, providing a unified evaluation standard for the game, allowing for fair comparison of different users' guessing performance. The time comparison subunit compares actual guessing times with the standard guessing time, providing an intuitive understanding of whether a user's guessing performance meets expectations. The efficiency evaluation subunit assesses guessing efficiency to understand a user's adaptability to game difficulty and their guessing skills. Based on the efficiency evaluation results, the game can provide personalized difficulty adjustments to enhance the user experience.

[0057] In particular, the difficulty analysis subunit analyzes the user's goal matching results to understand how the user adapts to the current difficulty level, thereby providing a more personalized gaming experience. The difficulty adjustment subunit dynamically adjusts the difficulty level based on the influence results of the difficulty analysis subunit, making the game gradually more challenging as the user progresses.

[0058] In particular, the information generation unit enhances game interactivity and user engagement by generating feedback related to game difficulty and user performance. The feedback recording unit, by recording feedback results, accumulates a large amount of user behavior data, providing a foundation for subsequent analysis and improvement. The information analysis unit, by analyzing user behavior patterns and difficulty adaptability, can gain in-depth user insights and help better understand the player base. The vocabulary update unit updates the vocabulary based on user analysis results, improving game adaptability and ensuring that game content matches user skills.

[0059] In particular, the Behavior Analysis subunit analyzes user behavior patterns to better understand their preferences and habits, thereby providing a more personalized gaming experience. The Difficulty Analysis subunit analyzes user adaptability to difficulty levels to provide appropriately challenging experiences, thereby promoting learning and skill improvement. The Result Synthesis subunit combines the results of behavioral and difficulty analyses to provide a more comprehensive and accurate user assessment, providing a basis for game design and adjustments.

[0060] In particular, the vocabulary screening subunit ensures that the vocabulary in the vocabulary library is more aligned with user interests and needs by selecting vocabulary based on user behavior patterns and adaptability analysis results. The difficulty adjustment subunit ensures that the game remains appropriately challenging for users by adjusting the difficulty level of vocabulary, avoiding being too easy or too difficult. The vocabulary verification subunit ensures that the updated vocabulary library is of high quality and suitable for the target user group through a verification process. BRIEF DESCRIPTION OF THE DRAWINGS

[0061] Figure 1 A schematic diagram of the structure of a fast picture-guessing word system based on artificial intelligence provided by an embodiment of the present invention;

[0062] Figure 2 A schematic diagram of the structure of a processing module in a fast-drawing word guessing system based on artificial intelligence provided by an embodiment of the present invention;

[0063] Figure 3 A schematic diagram of the structure of a control module in a fast-drawing word guessing system based on artificial intelligence provided by an embodiment of the present invention;

[0064] Figure 4 A schematic structural diagram of a feedback module in an artificial intelligence-based fast drawing word guessing system provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0065] In order to make the objects and advantages of the present invention more clearly understood, the present invention is further described below in conjunction with embodiments; it should be understood that the specific embodiments described herein are merely used to explain the present invention and are not intended to limit the present invention.

[0066] The preferred embodiments of the present invention are described below with reference to the accompanying drawings. It should be understood by those skilled in the art that these embodiments are only used to explain the technical principles of the present invention and are not intended to limit the scope of protection of the present invention.

[0067] It should be noted that, in the description of the present invention, terms such as "up", "down", "left", "right", "inside", and "outside" indicating directions or positional relationships are based on the directions or positional relationships shown in the accompanying drawings. This is only for the convenience of description and does not indicate or imply that the device or element must have a specific orientation, be constructed and operated in a specific orientation. Therefore, it cannot be understood as a limitation on the present invention.

[0068] Furthermore, it should be noted that, in the description of the present invention, unless otherwise expressly specified or limited, the terms "mounted," "connected," and "connected" should be understood in a broad sense. For example, they may refer to fixed connections, detachable connections, or integral connections; mechanical connections or electrical connections; direct connections or indirect connections through an intermediate medium; and internal communication between two components. Those skilled in the art will understand the specific meanings of the above terms in the present invention based on specific circumstances.

[0069] See also Figure 1 As shown, an embodiment of the present invention provides an artificial intelligence-based fast drawing word guessing system, comprising:

[0070] The acquisition module 10 is used to obtain the initial input information of each user in real time, including the initial word library type, game mode and difficulty level, and is also used to obtain the guessing input language and actual guessing time of each user in real time;

[0071] a processing module 20 connected to the acquisition module 10, configured to generate an initial image based on artificial intelligence technology and according to the initial vocabulary type, the difficulty level, and the game mode, and to analyze the semantic relevance between the initial image and the guessing input according to a predefined model to obtain an actual similarity match;

[0072] a control module 30 connected to the processing module 20, configured to compare the actual similarity matching degree with a preset standard similarity matching degree to obtain an initial matching result, determine a target matching result based on the initial matching result, the actual guessing time, and the preset standard guessing time, and adjust the difficulty level based on the target matching result and the game mode;

[0073] The feedback module 40 is connected to the control module 30, and is used to feed back the target matching result to the user, and provide different feedback information according to the game mode, record the feedback information to obtain a record result, analyze the record result to update the initial vocabulary type, and then obtain the target vocabulary type.

[0074] Specifically, the acquisition module acquires the user's initial input information in real time, including the user-selected initial vocabulary type (e.g., animals, food, celebrities), game mode (e.g., single-player, multiplayer, challenge mode), and difficulty level (e.g., easy, medium, hard). The user's guess input, i.e., the guessed word entered based on the displayed image, is acquired in real time. The user's actual guessing time, i.e., the time from viewing the image to entering the guessed word, is recorded. The processing module uses artificial intelligence technology to generate the corresponding initial image based on the user's selected initial vocabulary type and difficulty level. For example, an AI image generation algorithm (e.g., Stable Diffusion) is used to generate the image. In multiplayer competition mode, more challenging images may be generated. A predefined model (e.g., the CLIP model) is used to analyze the semantic relevance between the generated initial image and the user's guess input. An actual similarity match is calculated, i.e., the similarity score between the user's input word and the word represented by the image. The control module compares the actual similarity match obtained by the processing module with a preset standard similarity match. An initial matching result is determined to determine whether the user's guess is close to the standard answer. A target matching result is determined based on the initial matching result, the user's actual guessing time, and the preset standard guessing time. If the user takes too long to guess or the similarity is low, the difficulty level may be lowered; conversely, the difficulty level may be raised. The feedback module provides real-time feedback on target matching results to the user, for example, by displaying whether the guess was correct and the score on the screen. Taking the game mode into consideration, the target matching results are adjusted appropriately, such as introducing time penalties or additional scoring mechanisms in multiplayer competition modes. The difficulty level is adjusted based on the target matching results and game mode to ensure the game is challenging and fair. Target matching results are fed back to the user in an appropriate form (such as text, images, and sound). Different feedback information is provided based on the game mode, such as ranking and scoring information in multiplayer competition modes. The user's game results are recorded, including whether the guess was correct, time taken, and score. The recorded results are analyzed to identify user performance trends and preferences. Based on the analysis results, the initial vocabulary types are updated, removing words that are too easy or difficult for the user and adding new words. This leads to the optimization of the target vocabulary types, making the game more adaptable to the abilities and preferences of different users.

[0075] Specifically, the acquisition module obtains the user's guessing input and guessing time in real time, which enhances the interactivity and immediacy of the game and makes the user experience smoother and more interesting. The processing module uses artificial intelligence technology to generate images, ensuring the diversity and creativity of the images, and improving the playability and attractiveness of the game. By analyzing the semantic correlation between the image and the guessing input through a predefined model, it can provide a more accurate similarity match and ensure the fairness of the game results. The control module dynamically adjusts the difficulty level according to the user's actual performance (similarity match and guessing time) to ensure that the game is appropriately challenging for the user. The feedback module increases the interactivity of the game and the user's sense of participation by providing real-time feedback on the target matching results to the user.

[0076] Specifically, if Figure 2 As shown, the processing module 20 includes:

[0077] An image generating unit 21 generates an image matching the initial vocabulary type and the difficulty level using an artificial intelligence algorithm according to the initial vocabulary type, the game mode, and the difficulty level to obtain the initial image;

[0078] a feature extraction unit 22 connected to the image generation unit 21, configured to extract color, shape, and texture from the initial image to obtain a first feature result;

[0079] A semantic analysis unit 23, connected to the feature extraction unit 22, for converting the guessing word input into a semantic representation that matches the feature result to obtain a semantic result;

[0080] The similarity calculation unit 24 is connected to the semantic analysis unit 23 and is used to analyze the correlation between the feature result and the semantic result according to the predefined model to obtain the actual similarity matching degree.

[0081] Specifically, the image generation unit receives the initial vocabulary type, game mode, and difficulty level as input parameters. The initial vocabulary type is converted into a text description, which will guide the Stable Diffusion model to generate an image. The parameters of the Stable Diffusion model, such as the noise level and number of iterations, are adjusted according to the game mode and difficulty level to control the complexity and details of the image. The Stable Diffusion model is used to generate an image based on the text description, and the generated image is output as the initial image. The feature extraction unit receives the initial image from the image generation unit. The main color features of the image are extracted using a color recognition algorithm. Image processing techniques such as edge detection and contour analysis are applied to extract the shape features of the image. Texture analysis algorithms such as gray-level co-occurrence matrix (GLCM) are used to extract the texture features of the image. The color, shape, and texture features are integrated into a first feature result and output. The semantic analysis unit receives the guess word input by the user. The guess word is processed using NLP technology, performing operations such as word segmentation and stop word removal. The processed text is converted into a vector representation, and a pre-trained word embedding model can be used. The vector representation is output as a semantic result. The similarity calculation unit receives the first feature result from the feature extraction unit and the semantic result from the semantic analysis unit. The image encoder of the CLIP model is used to encode the initial image to obtain a feature representation of the image. The text encoder of the CLIP model is used to encode the user's guessed word (vector representation after semantic analysis) to obtain a feature representation of the text. The similarity between the image feature representation and the text feature representation is compared using cosine similarity or other similarity measurement methods. The calculated similarity score is output as the actual similarity match degree.

[0082] Specifically, the image generation unit generates images based on the initial vocabulary type, game mode, and difficulty level, creating customized image content that matches the game theme and difficulty level. The feature extraction unit converts image data into feature vectors, simplifying the data structure and facilitating computer processing and analysis. The semantic analysis unit uses semantic analysis to more accurately match user input with image content, improving the game's responsiveness. The similarity calculation unit analyzes the correlation between feature results and semantic results using a predefined model to quantify the degree of match between the user's guess and the image.

[0083] Specifically, the semantic unit includes:

[0084] A text preprocessing subunit, for cleaning and standardizing the guessing word input to obtain a processing result;

[0085] an embedding generation subunit, connected to the text preprocessing subunit, for converting the processing result into a vector representation to obtain a conversion result;

[0086] A feature extraction subunit, connected to the embedding generation subunit, for inputting the conversion result into a deep learning model to extract semantic features to obtain a second feature result;

[0087] A semantic generation subunit, connected to the feature extraction subunit, for aggregating the second feature result into a vector with a fixed length to obtain the semantic result.

[0088] Specifically, the text preprocessing subunit deletes special characters, punctuation marks, and redundant spaces in the input text. The input text is segmented into words or lexical units, which usually requires processing according to a specific language. The text is converted into a unified form, for example, converting all characters to lowercase or uppercase. Common, non-discriminative stop words, such as "de", "he", "shi", etc., are deleted. The embedding generation subunit uses a pre-trained word vector model, such as Word2Vec, GloVe, or BERT, to convert the segmented text into a vector representation. Each word is mapped to its corresponding vector, and these vectors capture the semantic information of the words. The feature extraction subunit selects an appropriate deep learning model to process semantic analysis, such as a convolutional neural network (CNN), a recurrent neural network (RNN), a long short-term memory network (LSTM), or a Transformer. The selected model is used to extract the deep features of the text, and these features can better represent the semantic content of the text. If a CNN is used, a pooling operation, such as max pooling, may be required on the feature map to obtain a feature vector with a fixed length. A comprehensive semantic vector is generated through the output layer of the model, and this vector represents the semantic content of the entire input text.

[0089] Specifically, the text preprocessing subunit reduces noise and improves the accuracy of subsequent processing steps by deleting irrelevant characters and stop words. By converting the text into a unified format, it ensures that the embedding generation subunit can process the input data consistently. The embedding generation subunit converts the text into vectors, enabling the computer to better understand and process natural language information. The feature extraction subunit can abstract the original text into a higher-level feature representation through the learning of the model, which is convenient for understanding and analysis. The semantic generation subunit realizes the compression and representation of the original text information through an aggregation operation, which is beneficial to improving computational efficiency and reducing storage costs.

[0090] Specifically, the similarity calculation unit includes:

[0091] A vector comparison subunit, for comparing the feature result and the semantic result to calculate the similarity between them to obtain a calculation result;

[0092] a similarity scoring subunit, connected to the vector comparison subunit, for assigning a similarity score to each pair of image features and text semantics according to the calculation results to obtain a plurality of score results;

[0093] The result aggregation subunit is connected to the similarity scoring subunit and is used to aggregate or average a plurality of the score results to obtain the actual similarity matching degree.

[0094] Specifically, the vector comparison subunit receives feature results (usually vector representations of image features) and semantic results (semantic vector representations of text). In order to ensure fairness of the comparison, the two vectors may need to be normalized to have unit length. The similarity between the two vectors is calculated using an appropriate similarity measurement method (such as cosine similarity, Euclidean distance, dot product, etc.). The similarity measurement result is output as the calculation result. The similarity scoring subunit assigns a similarity score to each pair of image features and text semantics based on the calculation results of the vector comparison subunit. If necessary, the score can be normalized to a specific range, such as 0 to 1. The assigned similarity score is output as the score result. The result aggregation subunit summarizes all similarity scores, which can be a simple summation or a more complex weighted average. If necessary, the average of all scores can be calculated as the final similarity score. The summarized or averaged score is output as the actual similarity match.

[0095] In this embodiment, assuming that both the image feature vector and the text semantic vector are two-dimensional, we will use cosine similarity to calculate the similarity between them.

[0096] Image feature vector (the result after feature extraction): image_feature = [2,3];

[0097] Text semantic vector (the result after semantic generation): text_semantic = [1, 2];

[0098] The calculation process of the vector comparison subunit: normalized vector (the vector here is already bounded, so it can be calculated directly):

[0099]

[0100] Calculate cosine similarity:

[0101]

[0102] In this example, there is only one similarity score, so there is no need to assign multiple scores. The similarity score is the cosine similarity value calculated above, which is 0.92.

[0103] Since there is only one score, the aggregation process directly uses this score. Therefore, the actual similarity match is 0.92.

[0104] Specifically, the vector comparison subunit calculates the similarity between the image feature vector and the text semantic vector, quantifying their proximity in vector space. The similarity calculation subunit uses scores to clearly visualize the degree of similarity between different images and texts, helping to distinguish between highly and less similar pairs.

[0105] Specifically, if Figure 3 As shown, the control module 30 includes:

[0106] A matching evaluation unit 31 is used to evaluate the actual similarity matching degree and to obtain the initial matching result;

[0107] A time evaluation unit 32 is used to compare the actual guessing time with the standard guessing time to evaluate the user's guessing efficiency and obtain a time evaluation result;

[0108] a result determination unit 33, connected to the matching evaluation unit 31 and the time evaluation unit 32, respectively, for determining the target matching result according to the initial matching result and the time evaluation result;

[0109] The difficulty adjustment unit 34 is connected to the result determination unit 33 and is used to adjust the difficulty level according to the target matching result and the game mode.

[0110] Specifically, the standard similarity match is a pre-set value that represents the level of word guessing accuracy that the system expects users to achieve. This value can be measured by having a representative group of users perform game tests and recording their average match score, which serves as a benchmark for setting the standard similarity match.

[0111] Specifically, the matching evaluation unit obtains the actual similarity matching degree and the standard similarity matching degree. The actual similarity matching degree is compared with the standard similarity matching degree. A matching degree threshold is set according to the game design requirements. If the actual similarity matching degree is higher than the threshold, the match is considered successful; otherwise, the match is considered failed. The result of successful or failed matching is output as the initial matching result. The time evaluation unit records the actual word guessing time and the standard word guessing time. By comparing the two, the word guessing efficiency is calculated, for example, by dividing the actual time by the standard time. Based on the efficiency value, the user's word guessing efficiency is evaluated. The evaluation result is output as the time evaluation result. The result determination unit obtains the initial matching result and the time evaluation result. Combining the two results, the final target matching result is determined. The target matching result is output to the received data. The difficulty adjustment unit obtains the target matching result and the current game mode. Based on the target matching result and the game mode, it decides whether to increase or decrease the game difficulty. According to the adjustment logic, the game difficulty level is modified. The new difficulty level is output.

[0112] Specifically, the match evaluation unit accurately assesses the accuracy of a user's guess by comparing the actual similarity match with the standard similarity match. The time evaluation unit measures the user's reaction speed and efficiency by comparing the actual guess time with the standard guess time. The result determination unit enables the game to dynamically adjust based on the user's overall performance to maintain its challenging and interesting nature. The difficulty adjustment unit dynamically adjusts the difficulty based on the user's gaming performance and gaming patterns to maintain the game's appeal.

[0113] Specifically, the time evaluation unit includes:

[0114] A time acquisition subunit, for acquiring or setting a standard guessing time for each word in the game to obtain the standard guessing time;

[0115] a time comparison subunit, connected to the time acquisition subunit, for comparing the actual word guessing time with the standard word guessing time to obtain a comparison result;

[0116] The efficiency evaluation subunit is connected to the time comparison subunit and is used to evaluate the word guessing efficiency according to the comparison result to obtain the time evaluation result.

[0117] Specifically, the time acquisition subunit collects historical word guessing data from the game, including users' average guessing times at different difficulty levels. Based on this collected data, a standard guessing time is set for each word or difficulty level. This can be determined using statistical methods (such as calculating an average) or an expert system. The set standard guessing time is stored in a database or memory for subsequent comparison. The standard guessing time is output to the time comparison subunit. The time comparison subunit receives the actual guessing time and the standard guessing time. The actual guessing time is compared with the standard guessing time and the time difference or ratio is calculated. A threshold is set for the time difference to determine whether the user's guessing time is faster, slower, or closer to the standard time. The time comparison results are output to the efficiency evaluation subunit. The efficiency evaluation subunit receives the comparison results output by the time comparison subunit. Based on the comparison results, the user's guessing efficiency is evaluated. For example, if the actual guessing time is less than or equal to the standard time, the user is considered efficient; if the actual guessing time is greater than the standard time, the user is considered inefficient. The evaluation results are classified into several levels, such as efficient, average, and inefficient. The evaluation results of the guessing efficiency are output to the result determination unit.

[0118] Specifically, the time acquisition subunit sets a standard guessing time for each word, providing a unified evaluation standard for the game, allowing for fair comparison of guessing performance across different users. The time comparison subunit compares actual guessing times with the standard guessing time, providing an intuitive understanding of whether a user's guessing performance meets expectations. The efficiency evaluation subunit assesses guessing efficiency to understand a user's adaptability to game difficulty and their guessing skills. Based on the efficiency evaluation results, the game can provide personalized difficulty adjustments to enhance the user experience.

[0119] Specifically, the difficulty adjustment unit includes:

[0120] a difficulty analysis subunit, configured to analyze the target matching result and the impact of the game mode on the difficulty level to obtain an impact result;

[0121] The difficulty adjustment subunit is connected to the difficulty analysis subunit and is used to adjust the difficulty level according to the impact result.

[0122] Specifically, the difficulty analysis subunit collects information on target matching results and the current game mode. Evaluate the accuracy of the target matching results, the number of successes, the number of failures, etc. to determine the user's adaptation to the current difficulty level. Analyze the impact of the game mode (such as single-player mode, multiplayer mode, etc.) on the difficulty level, because different game modes may require different difficulty settings. Based on the target matching results and the game mode, calculate an impact factor or score to indicate the need to adjust the difficulty level. Output the impact result to the difficulty adjustment subunit. The difficulty adjustment subunit receives the impact result from the difficulty analysis subunit. According to the game design, set the rules for difficulty adjustment. For example, if the impact result is lower than a certain threshold, reduce the difficulty; if it is higher than a certain threshold, increase the difficulty. According to the impact result and adjustment rules, adjust the current difficulty level and apply the adjusted difficulty level to the game.

[0123] In this embodiment, assume there is a word guessing game with five levels of difficulty. The user's performance at difficulty level 3 is as follows:

[0124] The user made 10 attempts at difficulty level 3, of which 6 were successful.

[0125] Analyze the user's success rate, which is 6 successes divided by 10 attempts, and the success rate is 60%. Based on the game mode (single-player mode), set the following rules:

[0126] If the success rate is less than 50%, the user is considered to be performing poorly and the difficulty needs to be lowered.

[0127] If the success rate is above 70%, the user is considered to have performed well and the difficulty can be increased.

[0128] In this example, the user's success rate is 60%, which is moderate, but close to the threshold we set for increasing difficulty. Therefore, we decide to increase the difficulty. Since the current difficulty level is 3, increasing it by one level brings the new difficulty level to 4. However, because the game's maximum difficulty level is 5, the difficulty won't exceed this upper limit. As a result, the user's difficulty level in the next round will be adjusted to 4, as their performance on difficulty level 3 indicates they're ready for a higher challenge.

[0129] Specifically, the difficulty analysis subunit analyzes the user's goal matching results to understand how the user adapts to the current difficulty level, thereby providing a more personalized gaming experience. The difficulty adjustment subunit dynamically adjusts the difficulty level based on the influence results of the difficulty analysis subunit, making the game gradually more challenging as the user progresses.

[0130] Specifically, if Figure 4 As shown, the feedback module 40 includes:

[0131] An information generating unit 41 is configured to generate corresponding feedback information according to the game difficulty and the target matching result to obtain a feedback result;

[0132] A feedback recording unit 42, connected to the information generating unit 41, is used to store the feedback result in a database or a log file to obtain a record result;

[0133] an information analysis unit 43 connected to the feedback recording unit 42, configured to analyze the user's behavior pattern and difficulty adaptability according to the recording result to obtain the analysis result;

[0134] The word library updating unit 44 is connected to the data analyzing unit 43 and is used to update the initial word library type according to the analysis result.

[0135] Specifically, the information generation unit obtains the current game difficulty and target matching results. Based on the game difficulty and target matching results, specific feedback information is generated, such as "Congratulations on completing the challenge!" or "Keep up the good work, you will do better next time!" The generated feedback information is output. The feedback recording unit receives the feedback results. The feedback results are stored in a database or log file. Ensure that the feedback results are correctly recorded and can be used for subsequent analysis. The information analysis unit extracts feedback records from the database or log file. By analyzing the user's feedback records, the user's behavior pattern and difficulty adaptability are identified. The analysis results are output for further processing. The vocabulary update unit receives the analysis results of the user's behavior pattern and difficulty adaptability from the information analysis unit. Based on the analysis results, the initial vocabulary type is adjusted, such as adding or reducing certain types of vocabulary. Ensure that the vocabulary update is implemented correctly.

[0136] Specifically, the information generation unit generates feedback related to game difficulty and user performance, enhancing game interactivity and user engagement. The feedback recording unit, by recording feedback results, accumulates a large amount of user behavior data, providing a foundation for subsequent analysis and improvement. The information analysis unit analyzes user behavior patterns and difficulty adaptability to gain in-depth user insights, helping to better understand the player base. The vocabulary update unit updates the vocabulary based on user analysis results, improving game adaptability and ensuring that game content matches user skills.

[0137] Specifically, the information analysis unit includes:

[0138] a behavior analysis subunit, configured to analyze a user's behavior pattern to obtain a first analysis result;

[0139] a difficulty analysis subunit, configured to analyze the user's adaptability to the difficulty level to obtain a second analysis result;

[0140] The result synthesis subunit is connected to the behavior analysis subunit and the difficulty analysis subunit respectively, and is used to synthesize the first analysis result and the second analysis result to obtain the analysis result.

[0141] Specifically, the behavior analysis subunit collects user behavior data in the game, including the number of guesses, success rate, common vocabulary types, game time, etc. Statistical methods or machine learning algorithms are used to analyze the data to identify user behavior patterns, such as a tendency to choose specific types of vocabulary, performance at a specific difficulty level, etc. The identified behavior patterns are summarized as the first analysis result. The difficulty analysis subunit obtains user performance data at different difficulty levels, including success rate, average word guessing time, etc. The data is analyzed to evaluate the user's adaptability to each difficulty level and determine whether the user performs well or struggles at a certain difficulty level. The adaptability assessment results are summarized as the second analysis result. The result synthesis subunit receives the first analysis result and the second analysis result and integrates the two results. Taking into account the user's behavior pattern and difficulty adaptability, a comprehensive analysis result is obtained. The comprehensive assessment result is output as the final analysis result.

[0142] In this embodiment, there is a simple word guessing game in which users guess words, and the game has different difficulty levels. The following is a specific numerical example to describe the workflow of the information analysis unit:

[0143] User ID: 12345; Number of guesses: 50; Success rate: 60% (i.e., 30 out of 50 successful guesses); Common vocabulary type: Noun (70% of guesses were nouns); Game duration: 5 hours. The user tends to guess nouns. The user's success rate is high, but there is still room for improvement. Output: First analysis result: User 12345's behavior pattern indicates a preference for guessing nouns, with a high overall success rate, but may need more practice with verbs and adjectives. User performance at different difficulty levels is collected: Easy: 80% success rate, average guessing time 20 seconds; Medium: 50% success rate, average guessing time 40 seconds; Hard: 30% success rate, average guessing time 60 seconds. Adaptability assessment: The user performs well on the Easy difficulty level, but performs averagely on the Medium and Hard difficulty levels. Output: Second analysis result: User 12345 adapts well to the Easy difficulty level, but may need more challenge and practice on the Medium and Hard difficulty levels. Result integration: First analysis result: The user prefers nouns and has a high overall success rate. Second analysis results: The user adapts well to the easy difficulty level, but performs averagely on the medium and hard difficulty levels. Overall assessment: User 12345 performs strongly on noun vocabulary, but may need more support on more difficult levels. We recommend providing this user with more medium-difficulty noun practice, as well as some verb and adjective practice to balance their skills.

[0144] Comprehensive analysis results show that users 12345 show a preference for nouns in the game and perform well on the easy difficulty level. To improve their performance on higher difficulty levels, we recommend adjusting the recommended vocabulary types and providing more practice opportunities on the medium difficulty level.

[0145] Specifically, the Behavior Analysis subunit analyzes user behavior patterns to better understand their preferences and habits, thereby providing a more personalized gaming experience. The Difficulty Analysis subunit analyzes user adaptability to difficulty levels to provide appropriately challenging gameplay, thereby promoting learning and skill improvement. The Result Synthesis subunit combines the results of behavioral and difficulty analyses to provide a more comprehensive and accurate user assessment, providing a basis for game design and adjustments.

[0146] Specifically, the vocabulary updating unit includes:

[0147] a vocabulary screening subunit, configured to screen out updated vocabulary based on the analysis result to obtain a screening result;

[0148] a difficulty adjustment subunit, connected to the vocabulary screening subunit, for adjusting the difficulty level of the screening result to obtain an adjusted result;

[0149] The vocabulary verification subunit is connected to the difficulty adjustment subunit and is used to verify whether the adjustment result is suitable for the user group.

[0150] Specifically, the vocabulary screening subunit receives the analysis results output by the information analysis unit. Based on the analysis results, the user's behavior pattern and difficulty adaptability are determined, and new vocabulary that meets the user's characteristics is screened out. For example, if the analysis results show that the user likes vocabulary on a specific topic, more vocabulary on that topic is screened out. The screened vocabulary list is output as the screening result. The difficulty adjustment subunit obtains the screening results output by the vocabulary screening subunit. Based on the user's difficulty adaptability analysis results, the difficulty level of the screened vocabulary is adjusted. This may include changing the commonness, frequency of use, spelling complexity, etc. of the vocabulary. The adjusted vocabulary list is output as the adjustment result. The vocabulary verification subunit obtains the adjustment results output by the difficulty adjustment subunit. Verify whether the adjustment results are suitable for the user group in the following ways:

[0151] User testing: Test the new vocabulary list with a group of target users and gather feedback.

[0152] Expert review: Invite language experts or education experts to review the vocabulary list.

[0153] Data analysis: Analyze users’ performance data in the test and evaluate whether the vocabulary difficulty is appropriate.

[0154] Based on the results of the verification process, determine whether the adjustment results are adopted or further adjustments are needed.

[0155] Specifically, the vocabulary screening subunit selects vocabulary based on user behavior patterns and adaptability analysis results, ensuring that the vocabulary in the vocabulary library is more in line with user interests and needs. The difficulty adjustment subunit adjusts the difficulty level of vocabulary to ensure that the game remains appropriately challenging for users, avoiding being too easy or too difficult. The vocabulary verification subunit ensures that the updated vocabulary library is of high quality and suitable for the target user group through a verification process.

[0156] Thus far, the technical solutions of the present invention have been described in conjunction with the preferred embodiments shown in the accompanying drawings. However, it will be readily understood by those skilled in the art that the scope of protection of the present invention is obviously not limited to these specific embodiments. Without departing from the principles of the present invention, those skilled in the art may make equivalent changes or substitutions to the relevant technical features, and the technical solutions after such changes or substitutions will fall within the scope of protection of the present invention.

[0157] The foregoing description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Those skilled in the art will readily appreciate that the present invention is susceptible to various modifications and variations. Any modifications, equivalent substitutions, or improvements made within the spirit and principles of the present invention are intended to be within the scope of protection of the present invention.

Claims

1. A fast picture-guessing system based on artificial intelligence, characterized by: include: The acquisition module is used to obtain the initial input information of each user in real time, including the initial word library type, game mode and difficulty level, and is also used to obtain the guessing input language and actual guessing time of each user in real time; a processing module, connected to the acquisition module, configured to generate an initial image based on artificial intelligence technology and according to the initial vocabulary type, the difficulty level, and the game mode, and to analyze the semantic relevance between the initial image and the guessing input language according to a predefined model to obtain an actual similarity match; a control module, connected to the processing module, configured to compare the actual similarity matching degree with a preset standard similarity matching degree to obtain an initial matching result, determine a target matching result based on the initial matching result, the actual guessing time, and the preset standard guessing time, and adjust the difficulty level based on the target matching result and the game mode; A feedback module is connected to the control module, and is used to feed back the target matching result to the user, and provide different feedback information according to the game mode, record the feedback information to obtain a record result, analyze the record result to update the initial vocabulary type, and then obtain the target vocabulary type.

2. The fast-drawing word guessing system based on artificial intelligence according to claim 1 is characterized in that: The processing module includes: an image generating unit, which generates an image matching the initial vocabulary type and the difficulty level using an artificial intelligence algorithm according to the initial vocabulary type, the game mode, and the difficulty level to obtain the initial image; a feature extraction unit, connected to the image generation unit, for extracting color, shape and texture from the initial image to obtain a first feature result; a semantic analysis unit connected to the feature extraction unit, configured to convert the guessing word input into a semantic representation that matches the feature result to obtain a semantic result; A similarity calculation unit is connected to the semantic analysis unit and is used to analyze the correlation between the feature result and the semantic result according to the predefined model to obtain the actual similarity matching degree.

3. The fast picture guessing system based on artificial intelligence according to claim 2 is characterized in that: The semantic unit includes: A text preprocessing subunit, for cleaning and standardizing the guessing word input to obtain a processing result; an embedding generation subunit, connected to the text preprocessing subunit, for converting the processing result into a vector representation to obtain a conversion result; a feature extraction subunit, connected to the embedding generation subunit, for inputting the conversion result into a deep learning model to extract semantic features to obtain a second feature result; The semantic generation subunit is connected to the feature extraction subunit and is used to aggregate the second feature result into a vector of a fixed length to obtain the semantic result.

4. The fast picture guessing system based on artificial intelligence according to claim 3 is characterized in that: The similarity calculation unit includes: A vector comparison subunit, configured to compare the feature result and the semantic result to calculate the similarity between them and obtain a calculation result; a similarity scoring subunit, connected to the vector comparison subunit, for assigning a similarity score to each pair of image features and text semantics according to the calculation results to obtain a plurality of score results; The result aggregation subunit is connected to the similarity scoring subunit and is used to aggregate or average a plurality of the score results to obtain the actual similarity matching degree.

5. The fast picture guessing system based on artificial intelligence according to claim 4 is characterized in that: The control module includes: a matching evaluation unit, configured to compare the actual similarity matching degree with the standard similarity matching degree to obtain the initial matching result; a time evaluation unit, configured to compare the actual guessing time with the standard guessing time to evaluate the user's guessing efficiency and thereby obtain a time evaluation result; a result determination unit, connected to the matching evaluation unit and the time evaluation unit respectively, for determining the target matching result according to the initial matching result and the time evaluation result; A difficulty adjustment unit is connected to the result determination unit and is used to adjust the difficulty level according to the target matching result and the game mode.

6. The fast picture guessing system based on artificial intelligence according to claim 5 is characterized in that: The time evaluation unit includes: A time acquisition subunit, for acquiring or setting a standard guessing time for each word in the game to obtain the standard guessing time; a time comparison subunit, connected to the time acquisition subunit, for comparing the actual word guessing time with the standard word guessing time to obtain a comparison result; The efficiency evaluation subunit is connected to the time comparison subunit and is used to evaluate the word guessing efficiency according to the comparison result to obtain the time evaluation result.

7. The fast picture guessing system based on artificial intelligence according to claim 6 is characterized in that: The difficulty adjustment unit includes: a difficulty analysis subunit, configured to analyze the target matching result and the impact of the game mode on the difficulty level to obtain an impact result; The difficulty adjustment subunit is connected to the difficulty analysis subunit and is used to adjust the difficulty level according to the impact result.

8. The fast-drawing word guessing system based on artificial intelligence according to claim 7 is characterized in that: The feedback module includes: An information generating unit, configured to generate corresponding feedback information according to the game difficulty and the target matching result to obtain a feedback result; A feedback recording unit, connected to the information generating unit, for storing the feedback result in a database or a log file to obtain a record result; an information analysis unit, connected to the feedback recording unit, for analyzing the user behavior pattern and difficulty adaptability according to the recording result to obtain the analysis result; The word library updating unit is connected to the data analyzing unit and is used to update the initial word library type according to the analysis result.

9. The fast-drawing word guessing system based on artificial intelligence according to claim 8 is characterized in that: The information analysis unit includes: a behavior analysis subunit, configured to analyze a user's behavior pattern to obtain a first analysis result; a difficulty analysis subunit, configured to analyze the user's adaptability to the difficulty level to obtain a second analysis result; The result synthesis subunit is connected to the behavior analysis subunit and the difficulty analysis subunit respectively, and is used to synthesize the first analysis result and the second analysis result to obtain the analysis result.

10. The fast drawing word guessing system based on artificial intelligence according to claim 9 is characterized in that: The vocabulary updating unit includes: a vocabulary screening subunit, configured to screen out updated vocabulary based on the analysis result to obtain a screening result; a difficulty adjustment subunit, connected to the vocabulary screening subunit, for adjusting the difficulty level of the screening result to obtain an adjusted result; The vocabulary verification subunit is connected to the difficulty adjustment subunit and is used to verify whether the adjustment result is suitable for the user group.

Images