Emotional score determination model construction and emotional score determination method, device and equipment

Abstract

The present invention relates to the field of computer vision technology, and discloses the construction of an emotion score determination model and an emotion score determination method, device and equipment. The model construction method constructs an emotion score determination model, which is used to characterize the correlation between each first category label score and each second category label score. The first classification method includes multiple first category labels, and the second classification method includes multiple second category labels. When subsequently determining the first category label score of an emotion image to be analyzed, the emotion score determination model can be used to determine the second category label and score corresponding to the emotion image to be analyzed, which solves the defect in the related art that the scoring result of determining the emotion score under the second classification method based on a scale is not accurate enough.

Description

Technical Field

[0001] The present invention relates to the field of computer vision technology, and in particular to the construction of a sentiment score determination model and a sentiment score determination method, device, and apparatus. Background Art

[0002] The Traditional Chinese Medicine (TCM) Seven Emotions Scale is a tool used to assess an individual's emotional state. Based on TCM emotional theory, it categorizes emotions into seven basic categories: joy, anger, worry, thought, sadness, fear, and shock. These scales can help doctors and researchers understand an individual's emotional state and intensity. For individuals, participating in the assessment can increase their understanding of their own emotions and help them manage their emotional state.

[0003] Currently, when using the Traditional Chinese Medicine Seven Emotions Scale to evaluate emotions, participants are typically required to manually fill out the scale, which is then analyzed to determine the participant's emotion type and score. However, this scale-based approach to determining a participant's emotional score has the following flaws: The scale relies on the participant's subjective responses, which can lead to inaccurate evaluation results due to potential bias in responses, memory distortion, or the influence of social expectations. In face-to-face or telephone interviews, participants may be subject to social response bias, meaning they may be more inclined to give responses that are perceived as correct or socially desirable, rather than the actual situation. Most scale-based testing methods rely on the participant's recollections and self-reports, making them unable to effectively observe and record non-verbal or non-subjective behaviors, attitudes, or emotions. Summary of the Invention

[0004] In view of this, the present invention provides an emotion scoring determination model construction and an emotion scoring determination method, device and equipment to solve the problem that the evaluation results of emotion scoring using the Traditional Chinese Medicine Seven Emotions Scale are not accurate enough.

[0005] In a first aspect, the present invention provides a method for constructing a mood score determination model, the method comprising: obtaining multiple mood sample images; classifying the multiple mood sample images according to a first classification method to obtain multiple sets, wherein each mood sample image in the same set corresponds to the same first category label; determining the first category score of each mood sample image in each set; in each set, classifying each mood sample image in the set according to a second classification method to obtain multiple subsets, wherein one set corresponds to multiple subsets, and each mood sample image in the same subset corresponds to the same second category label; determining the second category score corresponding to each mood sample image in the subset; constructing a mood score determination model based on the second category scores and the first category scores corresponding to the emotion sample images in each subset, the mood score determination model being used to characterize the correlation between each first category label score and each second category label score.

[0006] The method for constructing a mood score determination model provided by the present invention classifies multiple mood sample images according to a first classification method to obtain multiple sets, each mood sample image in the same set corresponds to the same first category label, and the first category score of each mood sample image in each set is determined; in each set, each mood sample image in the set is classified according to a second classification method to obtain multiple subsets, each mood sample image in the same subset corresponds to the same second category label, and the second category score of each mood sample image in each subset is determined, and a mood score determination model is constructed based on the first category score and the second category score corresponding to the mood sample image in each subset, and the mood score determination model is used to characterize the correlation relationship between each first category label score and each second category label score. Subsequently, when determining the first category label score of the mood image to be analyzed, the mood score determination model can be used to determine the second category label and score corresponding to the mood image to be analyzed, which solves the defect that the scoring result of determining the mood score under the second classification method based on the scale in the related art is not accurate enough.

[0007] In an optional embodiment, the step of constructing an emotion score determination model based on the second category scores and first category scores corresponding to the emotion sample images in each subset includes: determining the first category score sequence and the second category score sequence of the corresponding subset based on the second category scores and the first category label scores of the emotion sample images in each subset; calculating the Pearson coefficient between the first category score sequence and the second category score sequence corresponding to each subset; and constructing the emotion score determination model based on the Pearson coefficient between the first category score sequence and the second category score sequence corresponding to each subset.

[0008] The method provided in this optional embodiment constructs a sentiment score determination model through the first category score sequence and the second category score sequence corresponding to each subset, so that the model construction effect is better.

[0009] In an optional embodiment, the step of classifying multiple emotion sample images according to a first classification method to obtain multiple sets includes: inputting the multiple emotion sample images into a preset emotion analysis model, so that the preset emotion analysis model outputs a first category label for each emotion sample image; and determining multiple sets based on the first category label of each emotion sample image.

[0010] The method provided in this optional embodiment classifies each emotion sample image in the first classification mode based on a preset emotion analysis model, which can make the classification result more accurate and the classification efficiency higher.

[0011] In an optional embodiment, in each set, each emotion sample image in the set is classified according to the second classification method to obtain multiple subsets and second category scores corresponding to each image in the subset, including: identifying the annotation information of each emotion sample image in the set; determining the second category label of each emotion sample image in the corresponding set based on the annotation information of each emotion sample image in the set; and determining multiple subsets in each set based on the second category label of each emotion sample image in the set.

[0012] In the second aspect, an embodiment of the present invention provides a method for determining an emotion score, which also includes: obtaining multiple emotion images to be analyzed of a target tester; processing the multiple emotion images to be analyzed according to a first classification method to obtain a first category label and score for each image to be analyzed; inputting the first category label and score of each image to be analyzed into an emotion score determination model, so that the emotion score determination model outputs scores for different images to be analyzed corresponding to each second category label, and the emotion score determination model is constructed by the emotion score determination model construction method of the first aspect or any corresponding embodiment thereof; determining the emotion score of the target tester based on the scores for different images to be analyzed corresponding to each second category label.

[0013] The emotion score determination method provided by the present invention uses an emotion score determination model to determine the second category label and score corresponding to each emotion image to be analyzed, which solves the defect of inaccurate scoring results in the related art of determining the emotion score under the second classification method based on a scale.

[0014] In an optional embodiment, the step of obtaining multiple emotional images to be analyzed of the target tester includes: obtaining a facial visual image video of the target tester; and splitting the target facial visual image video frame by frame to obtain multiple emotional images to be analyzed.

[0015] In an optional embodiment, the step of determining the target tester's emotional score based on the scores of each second category label corresponding to different images to be analyzed includes: determining the distribution probability of each second category label based on the scores of each second category label corresponding to different images to be analyzed; and determining the target tester's emotional score based on the distribution probability of each second category label.

[0016] In a third aspect, the present invention provides a device for constructing a mood score determination model, which includes: a first acquisition module for acquiring multiple mood sample images; a first classification module for classifying the multiple mood sample images according to a first classification method to obtain multiple sets, wherein each mood sample image in the same set corresponds to the same first category label; a first determination module for determining the first category score of each mood sample image in each set; a second classification module for classifying each mood sample image in a set according to a second classification method in each set to obtain multiple subsets, wherein one set corresponds to multiple subsets, and each mood sample image in the same subset corresponds to the same second category label; a second determination module for determining the second category score corresponding to each mood sample image in the subset; a construction module for constructing a mood score determination model based on the second category scores and first category scores corresponding to the mood sample images in each subset, wherein the mood score determination model is used to characterize the correlation between each first category label score and each second category label score.

[0017] In an optional embodiment, the construction module includes: a first determination submodule, used to determine the first category score sequence and the second category score sequence of the corresponding subset based on the second category score and the first category label score of the emotion sample image in each subset; a calculation submodule, used to calculate the Pearson coefficient between the first category score sequence and the second category score sequence corresponding to each subset; and a construction submodule, used to construct an emotion score determination model based on the Pearson coefficient between the first category score sequence and the second category score sequence corresponding to each subset.

[0018] In an optional embodiment, the first classification module includes: a processing submodule, used to input multiple emotion sample images into a preset emotion analysis model, so that the preset emotion analysis model outputs a first category label for each emotion sample image; a second determination submodule, used to determine multiple sets based on the first category label of each emotion sample image.

[0019] In an optional embodiment, the second classification module includes: an identification submodule for identifying the labeling information of each emotion sample image in the set; a third determination submodule for determining the second category label of each emotion sample image in the corresponding set based on the labeling information of each emotion sample image in the set; and a fourth determination submodule for determining multiple subsets in each set based on the second category label of each emotion sample image in the set.

[0020] In a fourth aspect, the present invention provides an emotion score determination device, which includes: a second acquisition module for acquiring multiple emotion images to be analyzed of a target tester; a processing module for processing the multiple emotion images to be analyzed according to a first classification method to obtain a first category label and score for each image to be analyzed; a third determination module for inputting the first category label and score of each image to be analyzed into an emotion score determination model, so that the emotion score determination model outputs scores for different images to be analyzed corresponding to each second category label, and the emotion score determination model is constructed by the emotion score determination model construction method of the first aspect or any corresponding embodiment thereof; a fourth determination module for determining the emotion score of the target tester based on the scores for different images to be analyzed corresponding to each second category label.

[0021] In an optional embodiment, the second acquisition module includes: an acquisition submodule for acquiring a facial visual image video of a target tester; and a splitting submodule for splitting the target facial visual image video frame by frame to obtain a plurality of emotion images to be analyzed.

[0022] In a fifth aspect, the present invention provides a computer device comprising: a memory and a processor, the memory and the processor being communicatively connected to each other, computer instructions stored in the memory, and the processor executing the computer instructions to thereby execute the emotion score determination model construction method of the above-mentioned first aspect or any corresponding embodiment thereof, or execute the emotion score determination method of the above-mentioned second aspect or any corresponding embodiment thereof.

[0023] In a sixth aspect, the present invention provides a computer-readable storage medium having computer instructions stored thereon, the computer instructions being used to enable a computer to execute the emotion score determination model construction method of the above-mentioned first aspect or any corresponding embodiment thereof, or to execute the emotion score determination method of the above-mentioned second aspect or any corresponding embodiment thereof. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] In order to more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the specific embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0025] Figure 1 is a flow chart of a method for constructing a sentiment score determination model according to an embodiment of the present invention;

[0026] Figure 2 is a flow chart of another method for constructing a sentiment score determination model according to an embodiment of the present invention;

[0027] Figure 3 is a flow chart of a method for determining an emotion score according to an embodiment of the present invention;

[0028] Figure 4 is a flowchart of another method for determining an emotion score according to an embodiment of the present invention;

[0029] Figure 5 is a structural block diagram of a device for constructing a sentiment score determination model according to an embodiment of the present invention;

[0030] Figure 6 is a structural block diagram of a device for determining an emotion score according to an embodiment of the present invention;

[0031] Figure 7 Schematic diagram of the hardware structure of a computer device according to an embodiment of the present invention. DETAILED DESCRIPTION

[0032] To make the purpose, technical solutions, and advantages of the embodiments of the present invention more clear, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without making creative efforts shall fall within the scope of protection of the present invention.

[0033] In the prior art, using the Traditional Chinese Medicine Seven Emotions Scale to score emotions typically involves manually filling out the scale and analyzing the completed scale to determine the participant's emotion type and score. However, this scale-based method of determining the participant's emotion score is significantly influenced by the subject's subjective experience, resulting in inaccurate results.

[0034] In view of this, an embodiment of the present invention provides a method for constructing an emotion score determination model, which can be applied to a processor to realize the construction of an emotion score determination model. The method provided by the present invention classifies multiple emotion sample images according to a first classification method to obtain multiple sets, each emotion sample image in the same set corresponds to the same first category label, and determines the first category score of each emotion sample image in each set; in each set, classifies each emotion sample image in the set according to a second classification method to obtain multiple subsets, each emotion sample image in the same subset corresponds to the same second category label, and determines the second category score of each emotion sample image in each subset. Based on the first category score and the second category score corresponding to the emotion sample image in each subset, an emotion score determination model is constructed. The emotion score determination model is used to characterize the correlation between each first category label score and each second category label score. Subsequently, when determining the first category label score of the emotion image to be analyzed, the emotion score determination model can be used to determine the second category label and score corresponding to the emotion image to be analyzed, thereby solving the defect of inaccurate scoring results in the related art of determining the emotion score under the second classification method based on a scale.

[0035] According to an embodiment of the present invention, an embodiment of a method for constructing a sentiment score determination model is provided. It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions, and although a logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0036] In this embodiment, a method for constructing a sentiment score determination model is provided, which can be used in the above-mentioned processor. Figure 1 FIG. 1 is a flow chart of a method for constructing a sentiment score determination model according to an embodiment of the present invention. Figure 1 As shown, the process includes the following steps:

[0037] Step S101: Acquire multiple emotion sample images.

[0038] For example, the emotion sample images may include emotion sample images of different emotional states and expressions. In order to cover a variety of participants, the emotion sample images may include characters demonstrating different emotions, emotionally excited scenes, or emotional stills.

[0039] Step S102 : classifying the plurality of emotion sample images according to a first classification method to obtain a plurality of sets, wherein each emotion sample image in the same set corresponds to the same first category label.

[0040] Exemplarily, the first classification method can be any emotion classification method. In an embodiment of the present application, the first classification method may include but is not limited to a method of using mental health emotion analysis indicators to classify emotions. The first classification method includes multiple categories, and each category corresponds to a first category label; based on the first classification method, multiple emotion sample images are classified to obtain multiple sets, and the emotion sample images in each set correspond to the same first category label. The first category labels under the first classification method and the corresponding information are shown in Table 1 below.

[0041] Table 1

[0042]

[0043]

[0044] Step S103: determining the first category score of each emotion sample image in each set.

[0045] Exemplarily, the first category score is used to characterize the intensity of the corresponding category of emotions. In the embodiments of the present application, through non-contact real-time visual data acquisition technology, artificial intelligence AI deep learning technology, combined with psychology, physiology and other technologies, emotional changes are inferred by detecting head and neck movements and fluctuations in emotional sample images, thereby determining the emotional score of each emotional sample image.

[0046] Step S104 , in each set, classify each emotion sample image in the set according to the second classification method to obtain multiple subsets, wherein one set corresponds to multiple subsets, and each emotion sample image in the same subset corresponds to the same second category label.

[0047] For example, the second classification method may be an emotion classification method. The second classification method is different from the first classification method. In the embodiments of the present application, the second classification method may include, but is not limited to, an emotion classification method based on the Seven Emotions Assessment Scale of Traditional Chinese Medicine. Common assessment items of the Seven Emotions Assessment Scale of Traditional Chinese Medicine are as follows: a. Joy: reflects emotions such as happiness, joy, and optimism; b. Anger: includes emotions such as anger, irritation, and anxiety; c. Worry: includes emotions such as melancholy, sadness, and frustration; d. Thinking: includes emotions such as thinking, worry, and nervousness; e. Sadness: includes emotions such as sadness, grief, and loss; f. Fear: includes emotions such as fear, terror, and panic; g. Shock: includes emotions such as surprise, shock, and uneasiness.

[0048] Step S105 : determining the second category score corresponding to each emotion sample image in the subset.

[0049] Illustratively, in an embodiment of the present application, a pre-set scoring rule may be used to determine the second category score corresponding to each emotion sample image in the subset.

[0050] Step S106 , constructing an emotion score determination model based on the second category scores and the first category scores corresponding to the emotion sample images in each subset, wherein the emotion score determination model is used to characterize the correlation between each first category label score and each second category label score.

[0051] Exemplarily, based on the second category scores and first category scores corresponding to the emotion sample images in each subset, the correlation between the second category scores and the first category scores in each subset can be determined, and an emotion score determination model can be constructed based on the correlation.

[0052] The method for constructing a model for determining an emotion score provided in the present embodiment classifies a plurality of emotion sample images according to a first classification method to obtain a plurality of sets, wherein each emotion sample image in the same set corresponds to the same first category label, and the first category score of each emotion sample image in each set is determined; in each set, each emotion sample image in the set is classified according to a second classification method to obtain a plurality of subsets, wherein each emotion sample image in the same subset corresponds to the same second category label, and the second category score of each emotion sample image in each subset is determined; an emotion score determination model is constructed based on the first category score and the second category score corresponding to the emotion sample images in each subset; the emotion score determination model is used to characterize the correlation between each first category label score and each second category label score; subsequently, when determining the first category label score of the emotion image to be analyzed, the emotion score determination model can be used to determine the second category label and score corresponding to the emotion image to be analyzed, thereby solving the defect in the related art that the scoring result of determining the emotion score under the second classification method based on the scale is not accurate enough.

[0053] In this embodiment, a method for constructing a sentiment score determination model is provided, which can be used in the above-mentioned processor. Figure 2 FIG. 1 is a flow chart of a method for constructing a sentiment score determination model according to an embodiment of the present invention. Figure 2 As shown, the process includes the following steps:

[0054] Step S201: Acquire multiple emotion sample images. Figure 1 Step S101 of the illustrated embodiment will not be described in detail here.

[0055] Step S202: classify the plurality of emotion sample images according to the first classification method to obtain a plurality of sets, wherein each emotion sample image in the same set corresponds to the same first category label. Figure 1 Step S102 of the illustrated embodiment will not be described in detail here.

[0056] Specifically, the above step S202 includes:

[0057] In step S2021, multiple emotion sample images are input into a preset emotion analysis model, so that the preset emotion analysis model outputs a first category label for each emotion sample image. For example, in embodiments of the present application, the preset emotion analysis model may be constructed using computer vision technology, artificial intelligence, a deep learning model, a convolutional neural network, or a facial recognition algorithm. This model can classify the emotion sample images using the first classification method. Artificial intelligence (AI) refers to intelligence achieved through computers. It is a multidisciplinary field encompassing machine learning, pattern recognition, natural language processing, intelligent control, computer vision, and others. In theory and practice, AI attempts to simulate human intelligence and thought processes, enabling machines to understand, reason, learn, and make decisions like humans. AI systems can process large amounts of data and extract patterns and regularities from them through learning. They then use this information to autonomously make decisions or predict future situations. Artificial intelligence is now widely used, including speech recognition, image recognition, intelligent customer service, and automated control. Deep learning is a machine learning technology that uses structures and algorithms similar to those of human neural networks. Its purpose is to learn feature representations of data and perform classification or prediction. Deep learning models typically consist of multiple layers, each of which uses computation to learn higher-level, abstract representations of data. Backpropagation algorithms are typically used to train deep learning models. Through multiple iterations of training and continuous parameter fine-tuning, model performance gradually improves. Deep learning models, such as convolutional neural networks and recurrent neural networks, are used for emotion recognition and assessment. They classify and quantify emotions from visual data such as facial expressions and voices. These models can learn and infer emotional states through training on large-scale datasets. Computer vision is a technical field that studies how computers can acquire, understand, and interpret visual information by processing digital images or visual data. It encompasses multiple disciplines, including image processing, pattern recognition, machine learning, and artificial intelligence, aiming to enable computers to possess capabilities similar to those of the human visual system. Computer vision techniques can be used to analyze and recognize facial expressions, body language, and other visual data. This involves related technologies such as facial expression recognition and action recognition.

[0058] In the embodiment of the present application, emotion analysis uses non-contact real-time visual data acquisition technology, artificial intelligence AI deep learning technology, combined with psychology, physiology and other technologies to infer emotional changes by detecting head and neck movements and fluctuations. Human emotional (3D) head and neck movements and fluctuations are detected through three-dimensional control, and accumulated as frame differences in several visual frames. The vibration parameters (frequency and amplitude) of each element (pixel) of the visual image are calculated to measure the tiny movements, time and space fluctuations of the measured target, thereby providing a means of quantitatively measuring the emotions of the measured target.

[0059] First, a video image must be acquired and broken down into a series of consecutive frames. For each frame, information about head and neck movement and fluctuations can be obtained by calculating the differences between adjacent frames. The temporal and spatial fluctuations of each pixel are determined by comparing the pixel values ​​between adjacent frames, calculating the grayscale value differences or the Euclidean distance between pixels.

[0060] Secondly, these fluctuation parameters can be further analyzed to measure subtle movements and emotional changes in the target. This is achieved by calculating vibration parameters such as frequency and amplitude. Frequency indicates the speed of head and neck movements and fluctuations, while amplitude indicates their intensity or magnitude. By analyzing changes in these vibration parameters, the target's emotional state can be inferred.

[0061] By analyzing the pixel differences between adjacent frames, emotional state can be determined. When the object is moving rapidly or undergoing drastic changes, the pixel differences are larger, reflecting a higher emotional state. Conversely, when the object is relatively stable, the pixel differences are smaller, reflecting a lower emotional state.

[0062] Different emotional states result in distinct head and neck movements and fluctuation patterns, which in turn manifest as distinct changes in vibration parameters. By collecting and analyzing vibration parameter data, a correlation model can be established between emotional states and vibration parameter changes. This allows the target's psychological and emotional state to be calculated. By testing the subject for 30 to 60 seconds, relevant psychological and physiological indicators can be obtained, with a corresponding algorithm for each emotion.

[0063] For example, when studying a person's anger state, first, a set of sample data, including both angry and non-angry states, is collected and the corresponding vibration parameters are extracted. Then, this sample data is used to train an emotion classification model, mapping the vibration parameters to angry or non-angry states. Finally, the trained model is used to classify the newly collected vibration parameter data, calculating the probability distribution of anger states or the intensity of anger emotions. The target's anger state can then be determined based on the vibration parameters.

[0064] Step S2022: Determine multiple sets based on the first category label of each emotion sample image. For example, in this embodiment of the present application, based on the first category label of each emotion sample image, images belonging to the same first category are grouped as a set to obtain multiple sets.

[0065] Step S203: Determine the first category score of each emotion sample image in each set. Figure 1 Step S103 of the illustrated embodiment will not be described in detail here.

[0066] Step S204: In each set, classify each emotion sample image in the set according to the second classification method to obtain multiple subsets, wherein one set corresponds to multiple subsets, and each emotion sample image in the same subset corresponds to the same second category label. Figure 1 Step S104 of the illustrated embodiment will not be described in detail here.

[0067] Specifically, the above step S204 includes:

[0068] Step S2041: Identify the annotation information for each emotion sample image in the set. For example, in this embodiment, by annotating each emotion sample image, the primary emotion expressed in each visual is determined. The emotion labels provided by the Traditional Chinese Medicine Seven Emotions Rating Scale can be used to assign a corresponding emotion score to each visual.

[0069] Step S2042: Determine a second category label for each emotion sample image in the corresponding set based on the annotation information of each emotion sample image in the set. For example, in this embodiment of the present application, the processor identifies the annotation information of each emotion sample image and, based on the annotation information, determines a second category label and a corresponding second category score for each emotion sample image in each set.

[0070] Step S2043: Determine multiple subsets in each set based on the second category label of each emotion sample image in the set. For example, based on the second category label of each emotion sample image in each set, the emotion sample images in each set belonging to the same second category are taken as a subset.

[0071] Step S205: Determine the second category score corresponding to each emotion sample image in the subset. Figure 1 Step S105 of the illustrated embodiment will not be described in detail here.

[0072] Step S206: construct an emotion score determination model based on the second category scores and first category scores corresponding to the emotion sample images in each subset. The emotion score determination model is used to characterize the correlation between each first category label score and each second category label score. Figure 1 Step S106 of the illustrated embodiment will not be described in detail here.

[0073] Specifically, the above step S206 includes:

[0074] Step S2061 : determining a first category score sequence and a second category score sequence of the corresponding subset based on the second category scores and the first category label scores of the emotion sample images in each subset.

[0075] Exemplarily, a second category score sequence is generated based on the second category score corresponding to each emotion sample image in each subset, and a first category score sequence is generated based on the first category score of each emotion sample image in the corresponding subset.

[0076] Step S2062 , calculating the Pearson coefficient between the first category score sequence and the second category score sequence corresponding to each subset.

[0077] Exemplarily, the Pearson coefficient between the first category score sequence and the second category score sequence corresponding to each subset is calculated, and the calculated Pearson coefficient can be used to characterize the correlation between the first category score and the second category score; in an embodiment of the present application, the correlation between the mental health emotion analysis index and the traditional Chinese medicine seven emotions assessment index is shown in Table 2 below.

[0078] Table 2

[0079]

[0080]

[0081] Step S2063 : constructing a sentiment score determination model based on the Pearson coefficient between the first category score sequence and the second category score sequence corresponding to each subset.

[0082] In this embodiment, a method for determining an emotion score is provided, which can be used in the above-mentioned processor. Figure 3 is a flow chart of a method for determining an emotion score according to an embodiment of the present invention. Figure 3 As shown, the process includes the following steps:

[0083] Step S301: obtaining a plurality of emotion images to be analyzed of a target test subject.

[0084] For example, the target tester may be any tester whose emotions are to be assessed, and the plurality of images to be analyzed may be facial expression images of the target tester.

[0085] Step S302 : Process the plurality of emotion images to be analyzed according to a first classification method to obtain a first category label and score for each image to be analyzed.

[0086] Exemplarily, the first category label and score of each image to be analyzed are determined by a first classification method. In an embodiment of the present application, the first category label and score of each image to be analyzed can be determined by a preset sentiment analysis model.

[0087] In step S303, the first category label and score of each image to be analyzed are input into the emotion score determination model, so that the emotion score determination model outputs scores corresponding to the second category labels of different images to be analyzed. The emotion score determination model is constructed using the emotion score determination model construction method of the above embodiment.

[0088] Exemplarily, the first category label of each image to be analyzed is input into the emotion score determination model, and the model outputs the score of the image to be analyzed under each second category label.

[0089] Step S304 : determining the target test subject's emotion score based on the scores of the different images to be analyzed corresponding to the second category labels.

[0090] The emotion score determination method provided in this embodiment uses an emotion score determination model to determine the second category label and score corresponding to each emotion image to be analyzed, which solves the defect of inaccurate scoring results in the related art of determining the emotion score under the second classification method based on a scale.

[0091] In this embodiment, a method for determining an emotion score is provided, which can be used in the above-mentioned processor. Figure 4 is a flow chart of a method for determining an emotion score according to an embodiment of the present invention. Figure 4 As shown, the process includes the following steps:

[0092] Step S401: Obtain multiple emotion images to be analyzed from the target tester. Figure 1 Step S301 of the illustrated embodiment will not be described in detail here.

[0093] Specifically, the above step S401 includes:

[0094] Step S4011, obtaining a facial visual image video of the target tester.

[0095] Exemplarily, the target tester can be any tester who needs to undergo emotion detection. In the embodiment of the present application, the facial visual image video can be an image of the target tester's facial expression changes within 60 seconds.

[0096] Step S4012: split the target face visual image video frame by frame to obtain multiple emotion images to be analyzed.

[0097] Exemplarily, the facial visual image is split frame by frame to obtain multiple split video images, and the obtained multiple frames of video images are used as multiple emotion images to be analyzed.

[0098] Step S402: Process the plurality of emotion images to be analyzed according to the first classification method to obtain the first category label and score of each image to be analyzed. Figure 1 Step S302 of the illustrated embodiment will not be described in detail here.

[0099] Step S403: Input the first category label and score of each image to be analyzed into the emotion score determination model, so that the emotion score determination model outputs the score corresponding to each second category label of different images to be analyzed. The emotion score determination model is constructed using the emotion score determination model construction method of the above embodiment. For details, please refer to Figure 1 Step S303 of the illustrated embodiment will not be described in detail here.

[0100] Step S404: Determine the target tester's emotion score based on the scores of the second category labels corresponding to the different images to be analyzed. Figure 1 Step S304 of the illustrated embodiment will not be described in detail here.

[0101] Specifically, the above step S404 includes:

[0102] Step S4041 : determining the distribution probability of each second category label based on the scores of the second category labels corresponding to the different images to be analyzed.

[0103] Exemplarily, the distribution probability of each second category label in the plurality of images to be analyzed is determined based on the score of each second category label corresponding to each image to be analyzed.

[0104] Step S4042: Determine the target test subject's emotion score based on the distribution probability of each second category label.

[0105] Exemplarily, in an embodiment of the present application, based on the distribution probability of the second category labels in multiple images to be analyzed, the second category label with the largest distribution probability can be used as the second category label of the target tester, and the emotional score of the target tester can be determined based on the scoring data corresponding to the second category label with the largest distribution probability.

[0106] This embodiment also provides a device for constructing a sentiment score determination model, which is used to implement the above-mentioned embodiments and preferred embodiments. Details that have already been described will not be repeated. As used below, the term "module" may refer to a combination of software and / or hardware that implements a predetermined function. Although the devices described in the following embodiments are preferably implemented in software, implementation using hardware, or a combination of software and hardware, is also possible and contemplated.

[0107] This embodiment provides a device for constructing a sentiment score determination model. Figure 5 Shown, including:

[0108] A first acquisition module 501 is used to acquire a plurality of emotion sample images;

[0109] A first classification module 502 is configured to classify the plurality of emotion sample images according to a first classification method to obtain a plurality of sets, wherein each emotion sample image in the same set corresponds to the same first category label;

[0110] A first determination module 503 is configured to determine a first category score for each emotion sample image in each set;

[0111] A second classification module 504 is configured to classify, in each set, each emotion sample image in the set according to a second classification method to obtain multiple subsets, wherein one set corresponds to multiple subsets, and each emotion sample image in the same subset corresponds to the same second category label;

[0112] A second determination module 505 is used to determine a second category score corresponding to each emotion sample image in the subset;

[0113] A construction module 506 is used to construct an emotion score determination model based on the second category scores and the first category scores corresponding to the emotion sample images in each subset, and the emotion score determination model is used to characterize the correlation between each first category label score and each second category label score.

[0114] In some optional implementations, the building block 506 includes:

[0115] A first determination submodule is configured to determine a first category score sequence and a second category score sequence for a corresponding subset based on the second category score and the first category label score of the emotion sample images in each subset;

[0116] A calculation submodule, used to calculate the Pearson coefficient between the first category score sequence and the second category score sequence corresponding to each subset;

[0117] The submodule is constructed to construct a sentiment score determination model based on the Pearson coefficient between the first category score sequence and the second category score sequence corresponding to each subset.

[0118] In some optional embodiments, the first classification module includes:

[0119] a processing submodule, configured to input a plurality of emotion sample images into a preset emotion analysis model, so that the preset emotion analysis model outputs a first category label for each emotion sample image;

[0120] The second determination submodule is configured to determine a plurality of sets based on the first category label of each emotion sample image.

[0121] In some optional embodiments, the second classification module includes:

[0122] The recognition submodule is used to identify the annotation information of each emotion sample image in the collection;

[0123] A third determination submodule is configured to determine a second category label for each emotion sample image in the corresponding set based on the annotation information of each emotion sample image in the set;

[0124] The fourth determining submodule is configured to determine a plurality of subsets in each set based on the second category label of each emotion sample image in the set.

[0125] This embodiment provides a device for determining an emotion score. Figure 6 Shown, including:

[0126] The second acquisition module 601 is used to acquire multiple emotion images to be analyzed from the target tester;

[0127] A processing module 602 is configured to process the plurality of emotion images to be analyzed according to a first classification method to obtain a first category label and score for each image to be analyzed;

[0128] A third determination module 603 is configured to input the first category label and score of each image to be analyzed into a sentiment score determination model, so that the sentiment score determination model outputs scores corresponding to each second category label for different images to be analyzed. The sentiment score determination model is constructed using the sentiment score determination model construction method of the above embodiment;

[0129] The fourth determination module 604 is configured to determine the target test subject's emotion score based on the scores of the different images to be analyzed corresponding to the second category labels.

[0130] In some optional implementations, the second acquisition module 601 includes:

[0131] The acquisition submodule is used to obtain the face visual image video of the target tester;

[0132] The splitting submodule is used to split the target face visual image video frame by frame to obtain multiple emotional images to be analyzed.

[0133] In some optional implementations, the fourth determining module includes:

[0134] a fifth determination submodule, configured to determine a distribution probability of each second category label based on scores of different images to be analyzed corresponding to each second category label;

[0135] The sixth determination submodule is configured to determine the emotion score of the target test subject based on the distribution probability of each second category label.

[0136] The further functional description of each of the above modules and units is the same as that of the above corresponding embodiments and will not be repeated here.

[0137] The emotion score determination model construction device and the emotion score determination device in this embodiment are presented in the form of functional units, where the units refer to ASIC (Application Specific Integrated Circuit) circuits, processors and memories that execute one or more software or fixed programs, and / or other devices that can provide the above functions.

[0138] The embodiment of the present invention also provides a computer device having the above Figure 5 The sentiment score determination model building device shown, or having the above Figure 6 The emotion score determination device shown.

[0139] See also Figure 7 , Figure 7 is a structural diagram of a computer device provided by an optional embodiment of the present invention, such as Figure 7 As shown, the computer device includes: one or more processors 10, memory 20, and interfaces for connecting various components, including high-speed interfaces and low-speed interfaces. Various components utilize different buses to communicate with each other and can be installed on a common mainboard or installed in other ways as needed. The processor can process the instructions executed in the computer device, including instructions stored in the memory or on the memory to display the graphical information of the GUI on an external input / output device (such as, a display device coupled to the interface). In some optional embodiments, if necessary, multiple processors and / or multiple buses can be used together with multiple memories and multiple memories. Equally, multiple computer devices can be connected, and each device provides part of the necessary operations (for example, as a server array, a group of blade servers, or a multi-processor system). Figure 7 A processor 10 is taken as an example.

[0140] The processor 10 may be a central processing unit, a network processor, or a combination thereof. The processor 10 may further include a hardware chip. The hardware chip may be an application-specific integrated circuit, a programmable logic device, or a combination thereof. The programmable logic device may be a complex programmable logic device, a field programmable gate array, a general purpose array logic, or any combination thereof.

[0141] The memory 20 stores instructions that can be executed by at least one processor 10, so as to enable at least one processor 10 to execute the method shown in the above embodiment.

[0142] The memory 20 may include a program storage area and a data storage area, wherein the program storage area may store an operating system and application programs required for at least one function; the data storage area may store data created based on the use of the computer device, etc. In addition, the memory 20 may include a high-speed random access memory, and may also include a non-transient memory, such as at least one disk storage device, a flash memory device, or other non-transient solid-state storage device. In some optional embodiments, the memory 20 may optionally include a memory remotely located relative to the processor 10, and these remote memories may be connected to the computer device via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0143] The memory 20 may include a volatile memory, such as a random access memory; the memory may also include a non-volatile memory, such as a flash memory, a hard disk or a solid-state drive; the memory 20 may also include a combination of the above types of memory.

[0144] The computer device further includes a communication interface 30 for the computer device to communicate with other devices or a communication network.

[0145] The embodiment of the present invention also provides a computer-readable storage medium. The above-mentioned method according to the embodiment of the present invention can be implemented in hardware, firmware, or implemented as a computer code that can be recorded in a storage medium, or implemented as a computer code that is originally stored in a remote storage medium or a non-temporary machine-readable storage medium and downloaded through a network and will be stored in a local storage medium, so that the method described herein can be stored in such software processing on a storage medium using a general-purpose computer, a dedicated processor, or programmable or dedicated hardware. Among them, the storage medium can be a magnetic disk, an optical disk, a read-only storage memory, a random access memory, a flash memory, a hard disk or a solid-state drive, etc.; further, the storage medium can also include a combination of the above-mentioned types of memory. It can be understood that a computer, a processor, a microprocessor controller or programmable hardware includes a storage component that can store or receive software or computer code. When the software or computer code is accessed and executed by a computer, a processor or hardware, the method shown in the above embodiment is implemented.

[0146] Although the embodiments of the present invention have been described with reference to the accompanying drawings, those skilled in the art may make various modifications and variations without departing from the spirit and scope of the present invention. Such modifications and variations are all within the scope defined by the appended claims.

Claims

1. A method for constructing a sentiment score determination model, characterized in that: The method comprises: Obtain multiple emotion sample images; Classifying the plurality of emotion sample images according to a first classification method to obtain a plurality of sets, wherein each emotion sample image in the same set corresponds to the same first category label, wherein the first classification method is a method of performing emotion classification using a mental health emotion analysis indicator; Determine a first category score for each emotion sample image in each set; In each set, classifying each emotion sample image in the set according to a second classification method to obtain multiple subsets, wherein one set corresponds to multiple subsets, and each emotion sample image in the same subset corresponds to the same second category label, wherein the second classification method is an emotion classification method based on the seven emotions assessment in traditional Chinese medicine; Determine the second category score corresponding to each emotion sample image in the subset; Constructing an emotion score determination model based on the second category scores and the first category scores corresponding to the emotion sample images in each subset, wherein the emotion score determination model is used to characterize the correlation between each first category label score and each second category label score; The step of constructing an emotion score determination model according to the second category scores and the first category scores corresponding to the emotion sample images in each subset includes: Determine a first category score sequence and a second category score sequence for the corresponding subset based on the second category score and the first category label score of the emotion sample images in each subset; Calculate the Pearson coefficient between the first category score sequence and the second category score sequence corresponding to each subset; The emotion score determination model is constructed based on the Pearson coefficient between the first category score sequence and the second category score sequence corresponding to each subset.

2. The method according to claim 1, characterized in that The step of classifying the plurality of emotion sample images according to a first classification method to obtain a plurality of sets includes: Inputting the plurality of emotion sample images into a preset emotion analysis model, so that the preset emotion analysis model outputs a first category label for each emotion sample image; A plurality of sets are determined based on the first category label of each emotion sample image.

3. The method according to claim 1, characterized in that In each set, the steps of classifying each emotion sample image in the set according to the second classification method to obtain multiple subsets and second category scores corresponding to each image in the subset include: Identifying the annotation information of each emotion sample image in the set; Determining a second category label for each emotion sample image in the corresponding set based on the annotation information of each emotion sample image in the set; A plurality of subsets in each set are determined based on the second category label of each emotion sample image in the set.

4. A method for determining an emotion score, characterized in that: The method further comprises: Acquire multiple emotion images to be analyzed from the target tester; Processing the plurality of emotion images to be analyzed according to a first classification method to obtain a first category label and score for each image to be analyzed; inputting the first category label and score of each image to be analyzed into a sentiment score determination model, so that the sentiment score determination model outputs scores corresponding to each second category label for different images to be analyzed, the sentiment score determination model being constructed using the sentiment score determination model construction method according to any one of claims 1 to 3; The emotion score of the target tester is determined based on the scores of the second category labels corresponding to different images to be analyzed.

5. The method according to claim 4, characterized in that The step of obtaining a plurality of emotion images to be analyzed from the target tester includes: Obtaining a face visual image video of a target tester; The target face visual image video is split frame by frame to obtain the multiple emotion images to be analyzed.

6. The method according to claim 4, characterized in that The step of determining the target tester's emotion score based on the scores of the second category labels corresponding to different images to be analyzed includes: Determining the distribution probability of each second category label based on the scores of different images to be analyzed corresponding to each second category label; The emotion score of the target tester is determined based on the distribution probability of each second category label.

7. A device for constructing a sentiment score determination model, characterized in that: The device comprises: A first acquisition module is used to acquire multiple emotion sample images; A first classification module is configured to classify the plurality of emotion sample images according to a first classification method to obtain a plurality of sets, wherein each emotion sample image in the same set corresponds to the same first category label; A first determination module, configured to determine a first category score for each emotion sample image in each set; A second classification module is used to classify each emotion sample image in each set according to the second classification method to obtain multiple subsets, wherein one set corresponds to multiple subsets, and each emotion sample image in the same subset corresponds to the same second category label; A second determination module is used to determine the second category score corresponding to each emotion sample image in the subset; A construction module is used to construct an emotion score determination model based on the second category scores and the first category scores corresponding to the emotion sample images in each subset, wherein the emotion score determination model is used to characterize the correlation between each first category label score and each second category label score; The building blocks include: A first determination submodule is configured to determine a first category score sequence and a second category score sequence of a corresponding subset based on the second category score and the first category label score of the emotion sample images in each subset; A calculation submodule, used to calculate the Pearson coefficient between the first category score sequence and the second category score sequence corresponding to each subset; The construction submodule is used to construct the emotion score determination model based on the Pearson coefficient between the first category score sequence and the second category score sequence corresponding to each subset.

8. The device according to claim 7, characterized in that The first classification module includes: a processing submodule, configured to input the plurality of emotion sample images into a preset emotion analysis model, so that the preset emotion analysis model outputs a first category label for each emotion sample image; The second determining submodule is configured to determine a plurality of sets based on the first category label of each emotion sample image.

9. The device according to claim 7, characterized in that The second classification module includes: an identification submodule, configured to identify the annotation information of each emotion sample image in the set; A third determining submodule, configured to determine a second category label for each emotion sample image in the corresponding set based on the annotation information of each emotion sample image in the set; The fourth determining submodule is configured to determine a plurality of subsets in each set based on the second category label of each emotion sample image in the set.

10. A device for determining an emotion score, characterized in that: The device comprises: The second acquisition module is used to acquire multiple emotion images to be analyzed from the target tester; a processing module, configured to process the plurality of emotion images to be analyzed according to a first classification method to obtain a first category label and a score for each image to be analyzed; a third determination module, configured to input the first category label and score of each image to be analyzed into a sentiment score determination model, so that the sentiment score determination model outputs scores corresponding to the second category labels of different images to be analyzed, wherein the sentiment score determination model is constructed using the sentiment score determination model construction method according to any one of claims 1 to 3; The fourth determination module is configured to determine the emotion score of the target test subject based on the scores of the second category labels corresponding to the different images to be analyzed.

11. The device according to claim 10, characterized in that The second acquisition module includes: The acquisition submodule is used to obtain the face visual image video of the target tester; The splitting submodule is used to split the target face visual image video frame by frame to obtain the multiple emotion images to be analyzed.

12. A computer device, characterized in that: include: A memory and a processor, wherein the memory and the processor are communicatively connected to each other, the memory stores computer instructions, and the processor executes the emotion score determination model construction method according to any one of claims 1 to 3, or the emotion score determination method according to any one of claims 4 to 6 by executing the computer instructions.

13. A computer-readable storage medium, characterized in that The computer-readable storage medium stores computer instructions, which are used to enable a computer to execute the emotion score determination model construction method according to any one of claims 1 to 3, or the emotion score determination method according to any one of claims 4 to 6.

Images