A smart emotional interaction method of a smart table lamp, a smart table lamp, a product and a medium

Abstract

A smart emotional interaction method of a smart table lamp, a smart table lamp, a product and a medium. The method comprises: after selecting a corresponding fatigue threshold according to an instruction, acquiring eye, head and shoulder characteristics of a user; when the head change rate exceeds the threshold, it is judged whether the ratio of the shoulder and the head change rate is within a preset range, if the condition is met, a dynamic weighting factor is generated combined with the shoulder activity level, the dynamic environmental fatigue is calculated together with the eye and head fatigue, and a voice prompt is played when the threshold is exceeded; if the condition is not met or the head change rate does not exceed the threshold, the eye gaze duration and the head fixed posture duration are analyzed to calculate the weight factor and obtain the static environmental fatigue; when the static environmental fatigue exceeds the threshold, the light brightness is adjusted. The technical scheme provided by the application improves the accuracy of the smart table lamp in managing the fatigue state of the user in a complex scene.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of digital home intelligent terminals, and in particular to an intelligent emotional interaction method of a smart table lamp, a smart table lamp, a product and a medium. BACKGROUND

[0002] At present, with the rapid development of artificial intelligence and human-computer interaction technology, intelligent devices have been widely used in daily life and work. Especially in the scenes of office, study and other activities that require long-time concentration, intelligent devices can optimize user experience, improve work efficiency and life quality, and thus become very important tools.

[0003] In the related art, the smart table lamp usually collects the physiological characteristics of the user through the sensor to judge the fatigue degree. For example, based on the blinking frequency, head tilt amplitude and other parameters of the user to infer the fatigue degree of the user, and then prompts the user through prompt information. These methods can achieve a certain degree of fatigue monitoring and meet the user's regular needs in basic interaction functions.

[0004] However, the related art is prone to misjudgment in some special situations. For example, when the user is engaged in some activities with frequent head changes (such as manual operation), the user may be misjudged as being tired, but in fact it is not. In addition, such misjudgment may cause the smart table lamp to send a prompt signal when the user is not tired, which not only fails to provide effective help, but also may interfere with the user's normal work. The main reason for such misjudgment is that the parameters relied on lack accurate reflection of the actual state of the user in some special situations, resulting in insufficient accuracy of the fatigue assessment result in complex scenes. SUMMARY

[0005] The present application provides an intelligent emotional interaction method of a smart table lamp, a smart table lamp, a product and a medium, which is used to improve the accuracy of the smart table lamp in managing the fatigue state of the user in complex scenes.

[0006] In a first aspect of the present application, an intelligent emotional interaction method of a smart table lamp is provided, which comprises:

[0007] According to the user input instruction, a corresponding preset fatigue threshold is selected; the eye features, head features and shoulder features of the user are obtained; in the case that the change rate of the head features is greater than a preset head movement threshold, it is determined whether the ratio of the change rate of the shoulder features to the change rate of the head features is within a preset head-shoulder ratio range; if not, a dynamic weighting factor inversely proportional to the shoulder activity degree of the shoulder movement is generated, and a dynamic environmental fatigue is calculated by combining the eye fatigue and the head fatigue obtained by combining the eye features and the head features with the dynamic weighting factor; when the dynamic environmental fatigue exceeds the fatigue threshold, a voice prompt is played; if yes or in the case that the change rate of the head features is not greater than the preset head movement threshold, the gaze duration of the eye fixation target and the fixed posture duration of the head movement are obtained, and corresponding gaze weight factors and fixed posture weight factors are calculated respectively; a static environmental fatigue is calculated according to the eye fatigue, the head fatigue, the gaze weight factor and the fixed posture weight factor; when the static environmental fatigue exceeds the fatigue threshold, the light brightness is adjusted.

[0008] In the above embodiment, in a dynamic scene (such as manual operation), by detecting whether the head moves and the synchronization relationship between the head and the shoulder, if the head-shoulder movement is not synchronized, it is determined that the dynamic activity state. At this time, a dynamic weighting factor is generated according to the shoulder activity degree, and the more active the shoulder is, the lower the threshold of the head fatigue is. The dynamic environmental fatigue is calculated by combining the eye fatigue and the head fatigue, and if it exceeds the threshold, the user is reminded through the voice prompt. In a static scene (such as long-time office work or reading), if the head does not move or the head-shoulder is synchronized, it is determined that the static environment. At this time, the gaze duration and the fixed posture duration of the user are obtained, and the gaze weight factor and the fixed posture weight factor are calculated. The static environmental fatigue is calculated by combining the eye and head fatigue, and if it exceeds the threshold, the user is reminded by adjusting the light brightness. The above technical steps effectively avoid the misjudgment problem of related technologies in complex scenes. For example, in a dynamic activity scene, by dynamically analyzing the head-shoulder ratio and the shoulder activity degree, normal head movement is avoided from being misjudged as a fatigue state; and in a static scene, by introducing the gaze weight and the fixed posture weight, the visual and physical fatigue caused by the user's long-time fixed posture or gaze target is captured. This multi-dimensional parameter fusion method for fatigue evaluation in dynamic and static environments improves the accuracy of the intelligent table lamp in managing the user's fatigue state in complex scenes and avoids unnecessary interference.

[0009] In combination with some embodiments of the first aspect, in some embodiments, according to the shoulder activity degree of the shoulder movement, a dynamic weighting factor inversely proportional to the shoulder activity degree is generated, and a dynamic environmental fatigue is calculated by combining the eye fatigue and the head fatigue obtained by combining the eye features and the head features with the dynamic weighting factor, specifically including:

[0010] When multiple tasks are included in the instruction, the task types of the multiple tasks are acquired, and according to a preset task difference quantization rule, task bases corresponding to the task types are identified; the switching operation of the user between different tasks is monitored, the switching frequency at each time of switching tasks is recorded, and the task base before switching and the task base after switching corresponding to the tasks before and after switching are recorded, and the task base before switching and the task base after switching are subtracted to obtain a switching difference degree; the switching frequency and the multiple switching difference degrees are calculated according to a preset fatigue accumulation factor calculation rule to obtain a fatigue accumulation factor; the shoulder feature is quantized into a shoulder activity value; the shoulder activity value and the fatigue accumulation factor are processed in a nonlinear combination manner to generate a dynamic weighting factor of head fatigue; the head fatigue and the dynamic weighting factor are multiplied and added to the eye fatigue to calculate a dynamic environmental fatigue.

[0011] In the above embodiment, by acquiring the task types of the multiple tasks and calculating the task bases according to the preset task difference quantization rule, the complexity between different tasks is identified. The switching frequency of the user in the task switching and the difference between the task bases before and after switching are monitored, the switching difference degree is calculated, and the fatigue accumulation factor is generated by combining the switching frequency and the switching difference degree. In addition, the shoulder feature is quantized into an activity value and nonlinearly combined with the fatigue accumulation factor to form a dynamic weighting factor of head fatigue. Finally, by dynamically adjusting the head fatigue and superimposing the eye fatigue, the dynamic environmental fatigue is calculated, which can comprehensively consider the interactive influence of cognitive fatigue caused by task switching and physical activity, dynamically adapt to the fatigue characteristics in different task scenarios, avoid misjudgment caused by task complexity or frequent switching, and thus improve the accuracy of the intelligent table lamp in managing the fatigue state of the user in a complex and multi-task environment.

[0012] In combination with some embodiments of the first aspect, in some embodiments, the switching frequency and the multiple switching difference degrees are calculated according to a preset fatigue accumulation factor calculation rule to obtain a fatigue accumulation factor, specifically including:

[0013] When the task switching is monitored, the redirection duration of the user from the first task switching completion time to the second task is monitored; the ratio of the redirection duration to the preset task start duration is calculated to obtain a task switching coefficient; the switching frequency and the multiple switching difference degrees are calculated according to a preset fatigue accumulation factor calculation rule, and the result is multiplied by the task switching coefficient to obtain the fatigue accumulation factor.

[0014] In the above embodiment, when the user task switching is monitored, the time length from the user task switching from the first task to the redirection of the second task is recorded. Then, the task switching frequency, the difference between the tasks before and after the switching, and the redirection time length are all taken into account, so that a more accurate fatigue accumulation factor is obtained, which not only considers the frequency of task switching and the difference in task complexity, but also takes into account the additional cognitive burden caused by redirection in the switching process, reflecting the cumulative effect of task switching on fatigue. For example, when the user frequently switches between complex tasks and simple tasks, the task switching coefficient can capture the cognitive pressure of redirection, thereby avoiding misjudgment of the fatigue state due to high switching frequency alone. In this way, the fatigue state in a complex task environment can be more accurately evaluated, and the accuracy of the smart table lamp in managing the user fatigue in a multi-task scenario can be improved.

[0015] In combination with some embodiments of the first aspect, in some embodiments, the gaze time length of the eye fixation on the target and the fixed posture time length of the head action are obtained, and the corresponding gaze weight factor and fixed posture weight factor are calculated, specifically including:

[0016] The fixation duration of the user's eye fixation on the target and the fixation deviation frequency of the fixation direction are obtained, and the eye adjustment factor of the eye fatigue degree is calculated in combination with the blinking frequency and in combination with a preset adjustment factor calculation method. The head tilt angle of the head action and the head tilt change frequency of the head tilt angle are obtained, and the head adjustment factor of the head fatigue degree is calculated in combination with an adjustment factor calculation method. According to the ratio of the eye adjustment factor and the head adjustment factor, the gaze time length weight of the eye fixation on the target in the static environment and the fixed posture time length weight of the head action are adjusted to obtain the gaze weight factor and the fixed posture weight factor.

[0017] In the above embodiment, by quantifying the fatigue states of the eyes and the head and dynamically balancing the influence weights of the two, the fatigue characteristics of the user in a static scene can be reflected. For example, in a scene where the user fixes the screen for a long time but the head posture is relatively relaxed, the weight of the head action can be reduced, and the weight of the gaze time length is adjusted, focusing on the evaluation of visual fatigue. Through this multi-dimensional dynamic adjustment, the fatigue state of the user in a static environment can be more comprehensively captured, thereby improving the accuracy of the smart table lamp in managing the fatigue state of the user in a complex static scene.

[0018] In combination with some embodiments of the first aspect, in some embodiments, the static environment fatigue degree is calculated according to the eye fatigue degree, the head fatigue degree, the gaze weight factor, and the fixed posture weight factor, specifically including:

[0019] The gaze time coefficient and the head posture coefficient are calculated according to a difference between the gaze duration and a preset gaze time threshold and a difference between the head tilt angle and a preset head tilt angle threshold respectively; and the static environment fatigue degree is calculated according to the eye fatigue degree, the head fatigue degree, the gaze weight factor, the fixed posture weight factor, the gaze time coefficient and the head posture coefficient.

[0020] In the above embodiment, the gaze time coefficient is generated by calculating the difference between the gaze duration of the user and the preset gaze time threshold, and the head posture coefficient is generated by calculating the difference between the head tilt angle and the preset head tilt angle threshold. These two coefficients respectively quantify the degree of deviation of the user's visual and posture state in the static environment. In combination with the eye fatigue degree, the head fatigue degree, the gaze weight factor and the fixed posture weight factor, the static environment fatigue degree is further calculated. The introduction of the gaze time coefficient and the head posture coefficient makes the fatigue assessment more targeted. For example, in the long-time office scenario, the user may have a high degree of gaze concentration but less posture change, and the weights of various factors can be dynamically adjusted to reflect the fatigue state in the static environment. Through this multi-dimensional correlation analysis, the accuracy of the intelligent table lamp in managing the fatigue state of the user in the complex static environment is improved, and the fatigue prompting strategy is optimized to avoid misjudgment.

[0021] In combination with some embodiments of the first aspect, in some embodiments, when the dynamic environment fatigue degree exceeds the fatigue degree threshold, a voice prompt is played, specifically including:

[0022] When the dynamic environment fatigue degree exceeds the fatigue degree threshold, the operation pause points of the user are monitored; if an operation pause point with a duration exceeding a preset minimum pause time length is monitored within a preset short delay time window, a voice prompt is played at the end of the operation pause point; if no operation pause point is monitored within the short delay time window, and the dynamic environment fatigue degree continues to exceed the fatigue degree threshold, a preset buffer voice prompt tone is played after the short delay time window ends, and then a voice prompt is played.

[0023] In the above embodiment, when the dynamic environment fatigue exceeds the preset fatigue threshold, the operation pause point of the user is first monitored. If an operation pause point with a duration exceeding the minimum pause length is detected within the short delay time window, a voice prompt is played at the end of the pause point to avoid interrupting the smoothness of the user's operation. If no operation pause point meeting the conditions is detected, and the dynamic environment fatigue of the user continues to exceed the threshold, a buffer voice prompt sound is played at the end of the short delay time window to attract the user's attention, and then a voice prompt is played to ensure the effectiveness and timeliness of the prompt, which can dynamically adapt to the user's operation rhythm and avoid disturbing the user due to an untimely prompt in high-intensity or continuous operation. For example, when the user is performing a high-concentration task such as manual operation or game, the natural pause point of the user is preferred for prompting; and in a continuous and non-pause task, the user is gradually guided to pay attention to rest through the buffer prompt sound. This intelligent prompting strategy based on operation pause not only reduces invalid interference, but also improves the user's acceptance and efficiency of the prompt, thereby improving the accuracy of the smart table lamp in managing the user's fatigue state in complex dynamic scenarios.

[0024] In some embodiments of the first aspect, in some embodiments, the static environment fatigue is calculated according to the eye fatigue, the head fatigue, a gaze weight factor and a fixed posture weight factor, specifically comprising:

[0025] When the task type of the task currently performed by the user belongs to the preset high-precision visual task, the real-time color rendering index and the real-time color temperature value of the current lighting output are obtained, and are compared with the preset ideal color rendering index range and color temperature stability standard to calculate a lighting quality matching degree factor. The eye fatigue is corrected in combination with the lighting quality matching degree factor to obtain a corrected eye fatigue. The static environment fatigue is calculated according to the corrected eye fatigue, the head fatigue, the gaze weight factor and the fixed posture weight factor.

[0026] In the above embodiment, when it is detected that the task performed by the user belongs to a high-precision visual task, the real-time color rendering index and color temperature value of the current light are obtained, and are compared with the preset ideal color rendering index range and color temperature stability standard to calculate a light quality matching degree factor. The eye fatigue degree of the user is corrected through the factor, and the static environment fatigue degree is calculated in combination with the head fatigue degree, the gaze weight factor and the fixed posture weight factor. This method can dynamically reflect the actual influence of the lighting condition on the visual fatigue of the user, especially in high-precision visual tasks (such as image processing or medical image analysis), the instability of the light may significantly increase the visual burden of the user. By introducing the light quality matching degree factor, the fatigue caused by the lighting condition is quantified and corrected, so that the fatigue assessment is more close to the actual situation. For example, when the color rendering index deviates from the ideal range or the color temperature fluctuates greatly, the influence of the light on the fatigue is amplified, so as to prompt the user to optimize the lighting condition. Through the comprehensive analysis of the light and the fatigue, the complex high-precision task scene can be more scientifically adapted, and the accuracy of the management of the fatigue state of the user is improved.

[0027] In a second aspect, the embodiments of the present application provide a smart table lamp, comprising: one or more processors and a memory; the memory is coupled to the one or more processors, the memory is used to store computer program codes, the computer program codes comprise computer instructions, and the one or more processors invoke the computer instructions to enable the smart table lamp to perform the method described in the first aspect and any possible implementation manner of the first aspect.

[0028] In a third aspect, the embodiments of the present application provide a computer program product comprising instructions, which, when executed on a smart table lamp, cause the smart table lamp to perform the method described in the first aspect and any possible implementation manner of the first aspect.

[0029] In a fourth aspect, the embodiments of the present application provide a computer readable storage medium comprising instructions, which, when executed on a smart table lamp, cause the smart table lamp to perform the method described in the first aspect and any possible implementation manner of the first aspect.

[0030] It can be understood that the smart table lamp provided in the second aspect, the computer program product provided in the third aspect and the computer storage medium provided in the fourth aspect are all used to execute the intelligent emotional interaction method of the smart table lamp provided in the embodiments of the present application. Therefore, the beneficial effects that can be achieved are referred to the beneficial effects in the corresponding method, which will not be described here again.

[0031] The one or more technical solutions provided in the embodiments of the present application have at least the following technical effects or advantages:

[0032] 1. The present application avoids misjudging normal head movements as fatigue by dynamically analyzing head-to-shoulder ratio and shoulder activity in dynamic activity scenarios such as manual operation or exercise; and captures visual and physical fatigue caused by long-term fixed posture or gaze on the target by introducing gaze weight and fixed posture weight in static scenarios such as long-term office work or reading. This multi-dimensional parameter fusion fatigue evaluation method in dynamic and static environments improves the accuracy of the smart table lamp in managing user fatigue in complex scenarios and avoids unnecessary interference.

[0033] 2. The present application identifies the complexity between different tasks by obtaining the task types of multiple tasks and calculating the task base according to the preset task difference quantization rule. The switching frequency of the user in task switching and the difference between the task base before and after switching are monitored to calculate the switching difference degree, and the fatigue accumulation factor is generated by combining the switching frequency and the switching difference degree. In addition, the shoulder feature is quantized as an activity value and combined with the fatigue accumulation factor in a non-linear manner to form a dynamic weighting factor of head fatigue. Finally, by dynamically adjusting the head fatigue and superimposing the eye fatigue, the dynamic environment fatigue is calculated, which can consider the interactive influence of cognitive fatigue and physical activity caused by task switching, dynamically adapt to the fatigue characteristics in different task scenarios, avoid misjudgment caused by task complexity or frequent switching, and improve the accuracy of the smart table lamp in managing user fatigue in complex and multi-task environments.

[0034] 3. The present application records the duration from the completion of the first task switching to the redirection of the second task when the user task switching is monitored. Then, the task switching frequency, the difference between the tasks before and after switching, and the redirection duration are all added to the process to obtain a more accurate fatigue accumulation factor, which not only considers the frequency of task switching and the difference in task complexity, but also adds the additional cognitive burden caused by redirection during the switching process, reflecting the cumulative effect of task switching on fatigue. For example, when the user frequently switches between complex tasks and simple tasks, the task switching coefficient can capture the cognitive pressure of redirection, thereby avoiding misjudgment of fatigue status due to high switching frequency alone. In this way, the fatigue status in a complex task environment can be more accurately evaluated, and the accuracy of the smart table lamp in managing user fatigue in a multi-task scenario can be improved. BRIEF DESCRIPTION OF DRAWINGS

[0035] Figure 1 is a flowchart of the intelligent emotional interaction method of the smart table lamp in the embodiments of the present application;

[0036] Figure 2 is another flowchart of the intelligent emotional interaction method of the smart table lamp in the embodiments of the present application;

[0037] Figure 3is an exemplary hardware structure schematic diagram of the smart desk lamp in the embodiments of the present application. DETAILED DESCRIPTION

[0038] The terms used in the following embodiments of the present application are only for the purpose of describing specific embodiments, and are not intended to be limiting to the present application. As used in the specification and the appended claims of the present application, the singular forms "a," "an," and "the" are intended to include plural forms as well, unless the context clearly indicates otherwise. It will be further understood that the terms "and / or" as used herein refer to and encompass any or all possible combinations of one or more of the listed items.

[0039] Hereinafter, the terms "first", "second" are only for the purpose of description, and cannot be understood as implying or suggesting relative importance or implicitly indicating the number of the indicated technical features. Therefore, the features defined with "first", "second" can explicitly or implicitly include one or more of the features, and in the description of the embodiments of the present application, the meaning of "a plurality of" is two or more, unless otherwise specified.

[0040] In the related art, the smart desk lamp usually detects fatigue based on the physiological characteristic parameters of the user (such as blink frequency, head tilt angle, etc.), and reminds the user to pay attention to rest through a simple prompt function. These methods can achieve a certain degree of fatigue monitoring in a conventional context and meet the basic needs of users. However, in complex scenarios, these technologies are prone to misjudgment. For example, when the user is engaged in activities with frequent head movement or large body dynamic changes (such as sports or manual operation), normal dynamic behavior may be incorrectly judged as a fatigue state; while in a static scenario of long time stillness or screen fixation, the fatigue actually produced by the user due to long time fixed posture or visual concentration may be ignored. This misjudgment is mainly due to the lack of accurate reflection of the actual state of the user in certain special situations, resulting in insufficient accuracy of fatigue assessment results in complex scenarios.

[0041] In the embodiment of the present application, an intelligent emotional interaction method based on dynamic and static scene distinction is provided. First, the fatigue threshold is selected by user instruction for personalized setting. Then, when the head movement is detected, the sensitivity of fatigue evaluation is dynamically adjusted by combining the synchronicity and activity level of shoulder movement, so as to avoid misjudgment of normal activity as fatigue in a dynamic scene. In a static environment, the gaze duration and fixed posture duration are obtained, and the eye fatigue, head fatigue and weight factor are combined to capture the fatigue state caused by long-time fixed posture or gaze target. In addition, through multi-dimensional analysis of dynamic and static scenes, the present technology solves the recognition deviation problem of related technologies in complex situations. Whether it is dynamic activity or static task, the calculation weight of each parameter value in fatigue is changed in combination with the actual action of the user, and the calculated fatigue is more in line with the actual situation of the user, which improves the accuracy of the smart table lamp in managing the fatigue state of the user in complex scenes and avoids unnecessary interference.

[0042] Figure 1 is a flowchart of the intelligent emotional interaction method of the smart table lamp in the embodiment of the present application, which includes the following steps:

[0043] S101, selecting a corresponding preset fatigue threshold according to the instruction input by the user.

[0044] Specifically, the task instruction input by the user is received through the input module (such as a physical button, a voice input or a keyboard input) of the smart table lamp. The input module can be designed in multiple forms, for example, preset buttons such as “reading”, “computer”, “handwork” and the like are set on the table lamp, or the natural language instruction (such as “I want to do handwork”) of the user is received through the voice recognition module, or the text input of the user is accepted through the touch screen or the keyboard. According to the instruction input by the user, the matching threshold is found in the pre-stored task and fatigue threshold correspondence table. If the user inputs multiple tasks, the fatigue threshold values of multiple tasks are weighted and averaged or the average value of the fatigue threshold values corresponding to multiple tasks is directly calculated according to certain weight rules (such as task importance or user priority setting), so as to calculate a comprehensive fatigue threshold, thereby adapting to the actual needs of the user.

[0045] S102, obtaining the eye features, head features and shoulder features of the user.

[0046] Specifically, the dynamic feature data of the eyes, head and shoulders of the user is collected in real time through the video acquisition module (such as a high-definition camera). For the eye features, the eye blinking frequency and eyeball gaze area of the user are identified through image processing algorithms. The eye blinking frequency can be obtained by capturing the opening and closing actions of the eyelids and calculating the time interval; the eyeball gaze area is obtained by tracking the gaze direction in the camera picture, mapping the gaze point to the preset coordinate system in the field, and obtaining the gaze coordinate value of the user.

[0047] The head feature is obtained based on a pose estimation technique, and the motion change of the user's head is detected by a camera, including the tilt angle, rotation angle and motion frequency of the head. These data can be obtained by analyzing the relative position change of the face through a key point detection algorithm (such as face key point detection). The shoulder feature is obtained by capturing the motion trajectory of the shoulder region, identifying the displacement direction, motion amplitude and activity level of the shoulder, and converting these dynamic information into numerical representation.

[0048] S103, if the change rate of the head feature is greater than the preset head motion threshold, determining whether the ratio of the change rate of the shoulder feature to the change rate of the head feature is within a preset head-shoulder ratio range;

[0049] If not, the following step S104 is performed;

[0050] If yes or in the case where the change rate of the head feature is not greater than the preset head motion threshold, the following step S106 is performed.

[0051] Specifically, first, the change rate of the head feature (i.e. the change degree of the head motion in unit time) is calculated and compared with the preset head motion threshold. If the change rate of the head feature is greater than the threshold, it indicates that the user's head is active, and the next step is determined. Then, the change rate of the shoulder feature is calculated, and the ratio of the change rate of the shoulder feature to the change rate of the head feature is determined. It is determined whether the ratio is within a preset head-shoulder ratio range. The head-shoulder ratio range is a parameter for distinguishing whether the head and the shoulder are active synchronously. Within the range, it indicates that the motion frequency of the head and the shoulder is consistent (i.e. the head and the shoulder move together), and outside the range, it indicates that the motion frequency is inconsistent (i.e. the head moves but the shoulder does not, or the head does not move but the shoulder moves, or the head and the shoulder do not move together).

[0052] If the motion frequency of the head and the shoulder is inconsistent (the ratio is outside the head-shoulder ratio range), it is determined that the user may be in a special scenario that requires frequent head movement but is not tired (such as manual operation requiring frequent head movement), and step S104 is performed to handle such special cases. If the change rate of the head does not exceed the threshold (i.e. the head does not move) or the motion frequency of the head and the shoulder is consistent, it is determined to be another case, and step S106 is performed for processing.

[0053] S104, generating a dynamic weighting factor inversely proportional to the shoulder activity level according to the shoulder activity level of the shoulder motion, and calculating the dynamic environmental fatigue degree by combining the eye fatigue degree and the head fatigue degree obtained by the eye feature and the head feature with the dynamic weighting factor.

[0054] Specifically, the shoulder activity level is calculated by the user's shoulder movement. The shoulder activity level reflects the frequency and amplitude of shoulder movement in a unit of time. By collecting shoulder movement data (such as displacement, speed, etc.) and quantifying it, a numerical index is generated. The shoulder activity level is used to generate a dynamic weighting factor, which is inversely proportional to the shoulder activity level, i.e. the more active the shoulder, the smaller the weighting factor. The dynamic weighting factor is used to adjust the weight of the head fatigue in the overall fatigue assessment.

[0055] The eye fatigue is calculated based on eye features, mainly based on the blinking frequency and the change of the eye gaze area. For example, high blinking frequency or gazing at the same target for too long will increase the eye fatigue. The head fatigue is based on the analysis of head movement features, such as head tilt angle, movement frequency and amplitude. Long time maintaining the same posture will increase the head fatigue. By default, the weights of eye fatigue and head fatigue are the same.

[0056] In a dynamic environment, the shoulder activity level is used to adjust the weight of the head fatigue. If the shoulder movement is active, the weighting factor decreases, the weight of the head fatigue decreases, and the weight of the eye fatigue increases relatively. Finally, the dynamic environment fatigue is calculated by the following formula: dynamic environment fatigue = eye fatigue × (1-dynamic weighting factor) + head fatigue × dynamic weighting factor.

[0057] In actual scenarios, if the user is doing handcrafting, the shoulder moves frequently and the activity level is high. A lower dynamic weighting factor is generated by the activity level of shoulder movement, reducing the weight of head fatigue. This is because frequent shoulder movement may lead to increased head movement, but it is not caused by fatigue, so reducing the weight of head fatigue can avoid misjudgment.

[0058] S105, when the dynamic environment fatigue exceeds the fatigue threshold, a voice prompt is played.

[0059] Specifically, after calculating the dynamic environment fatigue, the value is compared with the fatigue threshold. The fatigue threshold is a value preset according to the task type input by the user, which is used to determine whether the user's fatigue state reaches the degree that needs to be reminded. If the dynamic environment fatigue exceeds the threshold, the voice prompt function is triggered to remind the user that he or she may have entered a state of fatigue.

[0060] The implementation of voice prompt depends on the voice module inside the desk lamp, including the storage and playback device of prompt voice file. When the trigger condition is met, the specified audio file is called through the built-in algorithm, and the prompt content is played through the speaker. These prompt contents can include "please rest" "suggest you adjust your posture" or "you have worked for a long time" and other personalized voice.

[0061] In some embodiments of the present application, when the dynamic environmental fatigue exceeds the fatigue threshold, the operation pause points of the user can be monitored, and the voice prompt can be played at an appropriate time, so as to avoid interrupting the key operation or thinking process of the user, and improve the user experience and acceptance of the voice prompt.

[0062] When it is detected that the dynamic environmental fatigue first exceeds the fatigue threshold, the voice prompt is not immediately played, but enters a voice prompt preparation state. At this time, the operation behavior of the user is monitored to identify possible operation pause points. The prompt is avoided from being directly played when the user is focused on a high-frequency task or is in a thinking arrangement period, causing unnecessary interference.

[0063] In the voice prompt preparation state, the operation micro-rhythm of the user, including the blinking frequency and the change of head movement, is monitored by various sensors (such as a camera or an inertial measurement unit) and behavior analysis algorithms. Blink pause time: when the blinking frequency of the user pauses for more than a preset blinking pause time threshold, it is considered that the user may be in a short operation pause state. Head movement no change time: when the head movement of the user does not change for more than a preset head pause time threshold, it is considered as an operation pause point. These operation pause points usually occur when the user completes a task sub-stage or enters a short thinking arrangement. This monitoring method can intelligently capture the natural rhythm of the user's operation, and avoid playing the voice prompt at the key operation node or the short thinking focus point.

[0064] If an operation pause point with a duration exceeding a preset minimum pause length is monitored within a preset short delay time window, the voice prompt is selected to be played within a preset extremely short time after the operation pause point ends. The prompt time is avoided as much as possible from the key operation or thinking activity of the user, which enhances the naturalness and acceptance of the prompt.

[0065] If a suitable operation pause point cannot be monitored within the short delay time window, and the dynamic environmental fatigue continues to exceed the threshold, the voice prompt will be played after the time window ends. In this case, in order to reduce the interference to the user, a soft buffer prompt sound is played first, and then the voice prompt is played at a lower volume and speed. The buffer design can reduce the impact of the sudden prompt on the user, and improve the acceptance of the prompt.

[0066] By combining the recognition of user operation intermittent points and the design of short delay time window, the playing time of voice prompt can be intelligently judged to avoid user discomfort or key operation interruption caused by sudden prompt. This method is especially suitable for scenarios with high user cognitive load in dynamic environment, such as long-time high-frequency task switching or complex task thinking arrangement stage, which can improve user experience, and further improve the accuracy of intelligent table lamp in managing user fatigue state in complex scenarios, especially in dynamic environment with high-frequency task switching or long-time operation.

[0067] S106, the gaze duration of the eye fixation target and the fixation posture duration of the head action are obtained, and the corresponding gaze weight factor and fixation posture weight factor are calculated respectively.

[0068] Specifically, the eye gaze direction of the user is captured in real time by the camera and image processing algorithm, and it is judged whether the eye is always gazing at the same target. If the change amplitude of the gaze direction is within a certain threshold (such as the change of the gaze point coordinates is less than a preset value), it is determined that the user is gazing at the fixed target, and the timing starts to record the duration of the gaze. The gaze weight factor is proportional to the gaze duration, that is, the longer the gaze time, the higher the weight. Because long-time fixation on the fixed target will increase the eye fatigue, the increase of the weight factor can reflect this fatigue trend.

[0069] By monitoring the head action (such as the tilt angle, the rotation angle), it is judged whether the user's head remains stationary. If the head action amplitude is lower than the preset threshold, it is determined that the head is in the fixed posture state, and the timing starts to record the duration of the fixed posture. The fixed posture weight factor is proportional to the fixed posture duration, that is, the longer the fixed time, the higher the weight. This is because long-time maintenance of the head fixed posture is easy to cause muscle fatigue, and the increase of the fatigue degree weight can better reflect this state.

[0070] By calculating the gaze weight factor and the fixed posture weight factor respectively, the fatigue state of the user can be more comprehensively evaluated. In this case, the fatigue trend of the eye and the head is the main influencing factor, so these weight factors directly participate in the subsequent static environment fatigue degree calculation.

[0071] In some embodiments, by calculating the user eye adjustment factor and the head adjustment factor, and dynamically adjusting the gaze duration weight and the fixed posture duration weight according to the ratio of the two, the fatigue state of the user can be more accurately reflected, thereby improving the accuracy of the intelligent table lamp in managing the fatigue state of the user in complex scenarios.

[0072] First, the eye fixation features of the user are captured through sensors and algorithms, including the duration of fixation on a target, the frequency of fixation direction shifts, and the changes in blink frequency. Fixation duration is used to measure whether the user has been focusing on a target for a long time, while fixation shift frequency reflects whether the user frequently switches fixation targets. These two factors combined can more comprehensively capture the user's visual behavior. In addition, changes in blink frequency, as one of the key biological signals of fatigue, can reflect the degree of eye strain of the user. For example, when the blink frequency decreases, it may indicate that the user's eye strain increases, leading to the accumulation of eye fatigue. By integrating these feature data with preset adjustment factor calculation methods (such as weighted average or nonlinear function), the eye adjustment factor is calculated. The advantage of this is that it quantifies multiple dimensions of visual features into one indicator, more accurately reflecting the user's eye fatigue state.

[0073] By monitoring the head movement features of the user, including the tilt angle of the head and its change frequency, and combining the overall stability of the user's body posture, the head adjustment factor is calculated. The head tilt angle reflects whether the user's current posture deviates from the normal range, while the tilt change frequency measures whether the user has frequent head movements. For posture stability, shoulder movement data is analyzed, for example, when the head and shoulders are consistently stable, it is considered that the user's head fatigue accumulates slowly, thereby reducing the value of the head adjustment factor. Through the comprehensive calculation of these features, the head adjustment factor can accurately reflect the degree of head fatigue of the user in a static environment. This method avoids the problem of misjudging subtle head movements as fatigue.

[0074] By calculating the ratio of the eye adjustment factor to the head adjustment factor, the gaze duration weight and the fixed posture duration weight in a static environment are dynamically adjusted. When the eye adjustment factor is high, it is considered that the user's visual fatigue is dominant, so the weight of gaze duration is increased; conversely, when the head adjustment factor is high, the system will increase the weight of the fixed posture duration. This dynamic adjustment mechanism is based on the real-time behavior characteristics of the user and can adapt to different scenarios of fatigue performance.

[0075] After dynamic adjustment, the gaze weight factor and the fixed posture weight factor are recalculated according to the adjusted gaze duration weight and fixed posture duration weight. This calculation method combines the multi-dimensional features of the eyes and head, and can more comprehensively capture the user's fatigue state. Compared to the traditional method of using only fixation duration and fixed posture duration as weight calculation basis, this method solves the problem of fatigue evaluation bias caused by single feature by introducing a dynamic adjustment mechanism.

[0076] The above steps can more accurately capture the contribution proportion of the visual and posture features of the user in fatigue accumulation by calculating the eye adjustment factor and the head adjustment factor and dynamically adjusting the weight according to the ratio of the two, optimize the fatigue evaluation method in a static environment, and improve the accuracy of the smart desk lamp in managing the fatigue state of the user in a complex scene, especially in the case of diversified user behavior characteristics.

[0077] S107, calculate a static environment fatigue degree according to the eye fatigue degree, the head fatigue degree, the gaze weight factor and the fixed posture weight factor.

[0078] Specifically, the calculation formula of the static environment fatigue degree can be simply represented as: static environment fatigue degree = (eye fatigue degree x gaze weight factor) + (head fatigue degree x fixed posture weight factor). Through this weighted calculation method, the fatigue degree of the eyes and the head can be dynamically adjusted according to the different gaze and fixed posture time, so as to be more in line with the actual fatigue state in a static environment.

[0079] In some embodiments, when calculating the static environment fatigue degree, the gaze time coefficient and the head posture coefficient can also be introduced to further optimize the fatigue calculation model in combination with the eye fatigue degree, the head fatigue degree and the weight factor, thereby improving the accuracy of the smart desk lamp in managing the fatigue state of the user in a complex scene.

[0080] First, the gaze time coefficient is calculated according to the difference between the gaze duration of the user and the preset gaze time threshold. Specifically, the gaze time threshold is a reference value set according to the general visual habits or task type of the user, for example, the gaze time threshold for a reading task can be longer, while the threshold for a browsing task can be shorter. When the gaze duration of the user exceeds the threshold, it is considered that the visual burden of the user increases, and therefore the gaze time coefficient increases with the increase of the difference. This design can capture the visual fatigue characteristics of the user due to long-time gaze on a fixed target, and solves the problem that a single eye fatigue degree cannot accurately quantify the influence of gaze behavior on fatigue accumulation.

[0081] The head posture coefficient is calculated by the difference between the head tilt angle of the user and the preset head tilt angle threshold. The head tilt angle threshold is used to represent the head angle range of the user in a normal comfortable posture, for example, the head tilt angle threshold under a sitting posture task can be smaller, while some special tasks (such as drawing) can allow a larger angle range. When the head tilt angle of the user deviates from the threshold range by a large margin, the head posture coefficient increases accordingly, indicating that the head posture of the user can cause fatigue accumulation to accelerate. Through this coefficient, the influence of abnormal head posture on fatigue can be more accurately quantified.

[0082] In calculating the static environment fatigue, the eye fatigue and the head fatigue are combined with the gaze weight factor and the fixed posture weight factor respectively, and the gaze time coefficient and the head posture coefficient are introduced to modify the fatigue. Specifically, the calculation formula of the static environment fatigue can be expressed as: static environment fatigue = (eye fatigue x gaze weight factor x gaze time coefficient) + (head fatigue x fixed posture weight factor x head posture coefficient). By introducing the gaze time coefficient and the head posture coefficient, the influence of the fatigue characteristics can be dynamically enhanced or weakened, and the evaluation deviation caused by the change of a single feature can be avoided. For example, the head posture of the user may deviate to a certain extent but is not maintained for a long time, at which time the head posture coefficient will limit the over-high accumulation of the head fatigue, and the accuracy of the calculation result is maintained.

[0083] In the above steps, by introducing the gaze time coefficient and the head posture coefficient, the visual and posture characteristics of the user can be more comprehensively integrated, and the calculation model of the static environment fatigue is dynamically optimized. This method solves the problem of inaccurate evaluation of a single feature, improves the accuracy of the intelligent table lamp in managing the fatigue state of the user in a complex scene, and is particularly outstanding in the case of diversified accumulation of fatigue characteristics of the user.

[0084] S108, when the static environment fatigue exceeds the fatigue threshold, the light brightness is adjusted.

[0085] Specifically, when it is detected that the static environment fatigue exceeds the fatigue threshold, the light brightness adjustment mechanism is triggered. The adjustment of the light brightness is proportional to the static environment fatigue, that is, the higher the fatigue, the higher the light brightness. This design is based on research showing that appropriately increasing the environmental brightness can effectively alleviate visual fatigue and attention decline, and also improve the mental state of the user.

[0086] In actual operation, the precise adjustment of the light brightness is realized through a light control module (such as a PWM dimming technology or a smart light source driving chip). The static environment fatigue is mapped to a numerical interval in the light brightness range, and the corresponding brightness adjustment formula can be: light brightness = basic brightness + (static environment fatigue x adjustment coefficient). In this case, when the fatigue increases, the light brightness will gradually increase.

[0087] The parameters can also be adjusted according to the individual needs of the user, for example, the user is allowed to set the upper and lower limits of the light brightness to avoid uncomfortable light for the user. In addition, in order to avoid frequent fluctuations, the time interval of the brightness adjustment or the smooth transition mechanism can be set to make the change of the light brightness more natural.

[0088] In the above embodiment, by analyzing the dynamic activity in the dynamic activity scene (such as manual operation or exercise), the normal head movement is avoided to be misjudged as the fatigue state through the dynamic analysis of the head-shoulder ratio and the shoulder activity; and in the static scene (such as long-time office work or reading), the visual and physical fatigue caused by the long-time fixed posture or the gaze target is captured through the introduction of the gaze weight and the fixed posture weight. The fatigue evaluation method of the multi-dimensional parameter fusion in the dynamic and static environment improves the accuracy of the intelligent table lamp in the complex scene for the user fatigue state management, and avoids unnecessary interference.

[0089] In some embodiments of the present application, when the user frequently performs multi-task switching, the fatigue accumulation may be underestimated or misjudged due to the difference in task complexity and frequent switching. The intelligent emotional interaction method of the intelligent table lamp provided by the present application can dynamically calculate the fatigue accumulation factor by recording the task switching frequency, the task base difference and the reorientation duration, and generate a dynamic weighting factor in combination with the shoulder activity to evaluate the fatigue state in the dynamic environment.

[0090] As shown in Figure 2 Another flowchart of the intelligent emotional interaction method of the intelligent table lamp provided by the embodiments of the present application is shown, which includes the following steps:

[0091] S201, selecting a corresponding preset fatigue threshold according to a user input instruction.

[0092] S202, acquiring eye features, head features and shoulder features of the user.

[0093] S203, in the case that the change rate of the head features is greater than a preset head movement threshold, determining whether the ratio of the change rate of the shoulder features to the change rate of the head features is within a preset head-shoulder ratio range.

[0094] If not, the following step S204 is performed;

[0095] If yes or in the case that the change rate of the head features is not greater than the preset head movement threshold, the following step is performed, and then the following step S213 is returned.

[0096] S204, when the instruction includes multiple tasks, acquiring task types of the multiple tasks, and identifying task bases corresponding to the task types according to a preset task difference quantization rule.

[0097] Specifically, when multiple tasks are included in the instructions, the multiple tasks are first extracted by the instruction parsing module, and the task content is classified. For example, tasks can include different categories such as reading, writing, drawing, programming, data analysis, etc. The determination of the task type is usually based on natural language processing technology or pre-defined task classification rules, such as identifying the specific type of task through keywords, operation types, or context semantic analysis.

[0098] The task base number is a quantitative indicator that reflects the basic requirements of different tasks on the cognitive load of the user. The setting of the task base number is usually based on task complexity, attention demand, physical involvement (such as whether hand-eye coordination is required), and the fatigue accumulation characteristics of the task. For example, the task base number of a reading task may be lower because its cognitive load on the user is relatively small, while the task base number of a programming task may be higher because it requires high-intensity logical thinking and attention concentration. The pre-setting of the task base number can be done through experimental data, user feedback, or expert definition.

[0099] The base number difference between tasks is calculated through the task difference quantification rule. This rule evaluates the cognitive load switching difficulty between different tasks according to the numerical difference of the task base number. The larger the task base number difference, the more cognitive adjustment is required for switching between tasks, and the faster the accumulated cognitive fatigue may be. For example, when switching from a low-base number task (such as simple document reading) to a high-base number task (such as complex programming), the user's cognitive fatigue will significantly increase. By identifying the task base number and the task difference, the system can predict the potential impact of task switching on the user's fatigue state.

[0100] The introduction of the task base number solves the problem of difficulty in quantifying the difference in cognitive load accumulation speed during task switching. Through a unified task base number, the differences between tasks can be quantified more scientifically, thereby providing precise support for fatigue management and intervention strategies.

[0101] S205, monitor the switching operation of the user between different tasks, record the switching frequency when switching tasks each time, and the task base number before switching and the task base number after switching corresponding to the tasks before and after switching, and obtain the switching difference degree by subtracting the task base number before switching from the task base number after switching.

[0102] Specifically, the user's behavior is monitored in real time through an image acquisition device (such as a camera) to capture the user's switching operation between different tasks. The behavior characteristics of the user (such as gaze direction, head movement, hand operation) are matched with the pre-input task instructions. When it is detected that the user's behavior switches from a mode corresponding to task one (such as reading a document) to a mode corresponding to task two (such as writing code), it is determined that a task switching occurs.

[0103] When each task switching occurs, the time point of switching is recorded, and the number of task switching occurring in a unit time is calculated, and the switching frequency is taken as an index for quantifying the rhythm of user task switching.

[0104] After detecting the task switching, the base of the task before switching and the base of the task after switching are extracted from the task base database. The task base is a numerical value preset according to the complexity, cognitive load and fatigue accumulation characteristics of the task. The difference between the base of the task before switching and the base of the task after switching is calculated, and the switching difference degree is calculated: switching difference degree = base of task after switching - base of task before switching. The calculation result can reflect the degree of change of cognitive load caused by task switching.

[0105] S206, when the task switching is monitored, the redirection duration of the user from the completion time of the first task switching to the second task is monitored.

[0106] Specifically, first, the completion time of the user switching from the first task is judged. The behavior analysis model is used to detect whether the degree of association between the user and the first task is lower than the preset association threshold. The degree of association is calculated according to the matching degree of the user's operation behavior (such as mouse click, keyboard input, head direction, hand movement, etc.) and the characteristics of the task. When the degree of association is lower than the preset value, it is determined that the user has stopped performing the first task, and the time is marked as the starting point of the redirection duration.

[0107] After recording the starting point, the behavior characteristics of the user are continuously monitored in real time, including operation mode, line of sight direction, task area interaction, etc. When the degree of association between the user's behavior and the second task is higher than the preset association threshold, it is determined that the user has successfully oriented to the second task, and the time is marked as the end point of the redirection duration. The degree of association is determined by the matching degree of the user's current behavior and the second task, for example, the user's mouse action is concentrated in the window corresponding to the second task or performs the operation mode specific to the task.

[0108] S207, calculate the ratio of the redirection duration to the preset task startup duration to obtain the task switching coefficient.

[0109] Specifically, the preset task startup duration refers to the average startup time preset for each type of task according to experimental data, user self-setting or industry standard. The startup duration of each task is determined by factors such as task complexity, cognitive load and user proficiency.

[0110] The redirection duration is compared with the preset task startup duration to calculate the task switching coefficient.

[0111] By calculating the task switching coefficient, the actual task switching time length of the user can be compared with the task-specific starting time length, so as to quantify the efficiency of task switching. Such comparison can reveal the performance difference of the user in different task switching, and provide a basis for further analysis of the user's cognitive load accumulation.

[0112] S208, calculate the switching frequency and the plurality of switching difference degrees according to the preset fatigue accumulation factor calculation rule, and multiply the result by the task switching coefficient to obtain the fatigue accumulation factor.

[0113] Specifically, according to the preset fatigue accumulation factor calculation rule, the switching frequency and the plurality of switching difference degrees are calculated. The calculation rule usually adopts a weighted average or a nonlinear function method to quantify the comprehensive influence of switching frequency and task complexity on fatigue accumulation. For example: weighted average method: fatigue accumulation factor = switching frequency x ∑ (switching difference degree x preset weight) / total task switching times; nonlinear function method: enhance the influence of high switching frequency or high switching difference degree on fatigue accumulation through exponential or logarithmic function, for example: fatigue accumulation factor = switching frequency x log (1 + ∑ switching difference degree).

[0114] The design of these rules is based on experimental data and theoretical models. The weighted average method assigns weights to different switching difference degrees, considering the different contributions of each switching to fatigue accumulation, making the calculation result more accurate and suitable for scenarios where the influence of task difference is balanced. The nonlinear function method amplifies the influence of high switching frequency or high switching difference degree through exponential or logarithmic function, more sensitively capturing the acceleration effect of fatigue accumulation under extreme switching conditions, suitable for complex scenarios with high intensity and multiple task switching, and can more realistically reflect the nonlinear accumulation characteristics of fatigue, ensuring that the fatigue accumulation factor accurately reflects the dynamic influence of task switching on the user's fatigue state.

[0115] S209, quantize the shoulder feature as a shoulder activity value.

[0116] Specifically, the key values of the shoulder feature are extracted for quantization: movement amplitude: the displacement of the shoulder in the vertical and horizontal directions, calculated by coordinate change. Action frequency: the number of shoulder actions per unit time, obtained by time series analysis of the repeated patterns of shoulder action signals. Action speed: the speed of shoulder movement, the calculation formula is speed = displacement / time, obtained by the displacement difference of consecutive frames. Angle change: the rotation and inclination angle of the shoulder joint, extracted by bone model matching.

[0117] Based on the extracted shoulder action features, a shoulder activity value is calculated according to a preset weight. The quantification formula is as follows: shoulder activity value = a x movement amplitude + b x action frequency + g x action speed + d x angle change. Wherein, a, b, g, d are weight coefficients set according to experimental data or task requirements, used to adjust the influence of different features on the overall activity value. Through this weighted comprehensive calculation, multiple shoulder action features can be integrated into a unified index to reflect the overall level of shoulder activity.

[0118] S210, the shoulder activity value and the fatigue accumulation factor are processed by a nonlinear combination method to generate a dynamic weighting factor of head fatigue.

[0119] Specifically, the processing method of nonlinear combination is usually realized by exponential function, logarithmic function or polynomial function formula. Dynamic weighting factor = a x exp(b x shoulder activity value) + g x log(1 + fatigue accumulation factor), wherein a, b, g are preset adjustment parameters for determining the weight of shoulder activity value and fatigue accumulation factor on the final result. The exponential function is used to amplify the high motion influence of the shoulder activity value, while the logarithmic function smooths the effect of high fatigue accumulation factor, preventing the over amplification of a single high value on the result. In this way, the system can flexibly adapt to the fatigue characteristics in different scenarios, capturing the nonlinear influence of shoulder action and task switching on head fatigue.

[0120] The nonlinear combination method can more flexibly adjust the influence weight of input parameters, and adapt to the nonlinear influence of different fatigue sources on head fatigue. For example, in the case of high shoulder activity, the exponential function can amplify its influence, while when the task switching frequency is high, the logarithmic function can smooth the effect of high fatigue accumulation factor. Compared with the linear combination method, the nonlinear method is more consistent with the complex characteristics of actual fatigue accumulation. Through the above calculation, the dynamic weighting factor as the core index of head fatigue can reflect the user's comprehensive fatigue state in real time, especially the combined effect of physical fatigue and cognitive fatigue.

[0121] S211, multiplying the head fatigue and the dynamic weighting factor and adding the eye fatigue to calculate the dynamic environmental fatigue.

[0122] Specifically, first, the head fatigue is calculated by analyzing the frequency, amplitude and posture change of the user's head action, and combining the shoulder activity value and the fatigue accumulation factor comprehensive model, reflecting the overall fatigue level of the user's head. The dynamic weighting factor is calculated by nonlinearly combining the shoulder activity value and the fatigue accumulation factor, dynamically adjusting the influence weight of shoulder action and task fatigue on head fatigue. The eye fatigue is calculated by monitoring the blink frequency, gaze duration and pupil change parameters, representing the visual fatigue degree of the user.

[0123] Dynamic environment fatigue = head fatigue * dynamic weighting factor + eye fatigue. The product of head fatigue and dynamic weighting factor amplifies or reduces the influence of head fatigue on overall fatigue, and the dynamic weighting factor dynamically adjusts the head fatigue according to the changes in user shoulder movement and task switching; and the eye fatigue is independently added by addition, to ensure that the contribution of visual fatigue is accurately reflected. This calculation method integrates physical fatigue (head, shoulders), cognitive fatigue (task switching) and visual fatigue (eyes) to generate an index that comprehensively reflects the user's fatigue state.

[0124] S212, when the dynamic environment fatigue exceeds the fatigue threshold, playing a voice prompt.

[0125] S213, obtaining the gaze duration of the eye fixation target and the fixed posture duration of the head movement, and calculating the corresponding gaze weight factor and fixed posture weight factor.

[0126] S214, when the task type of the task currently performed by the user belongs to the preset high-precision visual task, obtaining the real-time color rendering index and real-time color temperature value of the current light output, and comparing with the preset ideal color rendering index range and color temperature stability standard, to calculate the light quality matching degree factor.

[0127] Specifically, the task is directly determined to belong to a high-precision visual task through a user input instruction. For example, when the user explicitly declares through voice, text or application interface that the current task is image processing, video color adjustment, medical image diagnosis or other operations requiring high-precision visual recognition, the task type can be directly confirmed.

[0128] Once the task is confirmed to be a high-precision visual task, the real-time light output parameters of the smart table lamp are obtained, including the color rendering index and the color temperature value. The color rendering index reflects the ability of the light source to restore colors, and the color temperature value describes the color tone characteristics of the light source. The real-time color rendering index is compared with the ideal color rendering index range, and the real-time color temperature value is compared with the preset color temperature stability standard. By comprehensively comparing these results, the light quality matching degree factor is calculated, and the matching degree factor is inversely proportional to the deviation degree of the color rendering index and the color temperature value. The closer the light is to the ideal standard, the higher the matching degree factor. This factor is used to quantify the degree of support of the current light for high-precision visual tasks.

[0129] In some embodiments, in addition to the lighting parameter, the diameter change of the user's pupil and the response stability to weak light intensity disturbance are also monitored by the camera. Rapid expansion or contraction of the pupil reflects the user's visual accommodation burden, especially when the lighting is unstable. These data are quantified as a visual accommodation stability index for evaluating the user's visual nervous system's adaptability to the ambient light. Visual accommodation stability index = 1 / (pupil diameter change amplitude x response time), the higher the stability, the smaller the pupil change, the larger the index value. The lighting quality matching degree factor and the visual accommodation stability index are fused to calculate a color recognition fatigue correction coefficient = lighting quality matching degree factor x visual accommodation stability index. This coefficient is used to correct the basic eye fatigue degree to generate a fine visual task type eye fatigue degree, which more accurately reflects the user's visual fatigue state in high-precision visual tasks.

[0130] The corrected fine visual task type eye fatigue degree is replaced by the original eye fatigue degree, and combined with the head fatigue degree, the gaze weight factor and the fixed posture weight factor to calculate the final static environment fatigue degree.

[0131] S215, correct the eye fatigue degree in combination with the lighting quality matching degree factor to obtain a corrected eye fatigue degree.

[0132] Specifically, the correction process is realized by combining the lighting quality matching degree factor with the basic eye fatigue degree. Corrected eye fatigue degree = basic eye fatigue degree x lighting quality correction coefficient (1 / lighting quality matching degree factor). When the lighting quality matching degree factor is high, the correction coefficient is close to 1, and the corrected eye fatigue degree changes little; when the matching degree factor is low, the correction coefficient amplifies the basic eye fatigue degree, reflecting the aggravating effect of poor lighting on visual fatigue. In this way, the eye fatigue degree can be dynamically adjusted to better fit the user's actual visual burden.

[0133] S216, calculate the static environment fatigue degree according to the corrected eye fatigue degree, the head fatigue degree, the gaze weight factor and the fixed posture weight factor.

[0134] S217, adjust the light brightness when the static environment fatigue degree exceeds the fatigue threshold.

[0135] Steps S201-S203, S212-S213, S216-S217 are similar to steps S101-S103, S105-S108 in the embodiment shown in Figure 1 The steps S101-S103, S105-S108 in the embodiment shown in

[0136] In the above embodiments, by monitoring the user's operational behavior during multi-task switching, the task types of multiple tasks are first obtained, and the task base number of each task is calculated according to the task difference quantification rules. During task switching, the switching frequency and the difference in task base number before and after the switch are recorded to assess the degree of switching difference. Simultaneously, the reorientation time from the first task to the second task is monitored, and its ratio to the preset task start time is calculated to generate a task switching coefficient. Combining the switching frequency, the degree of switching difference, and the task switching coefficient, a fatigue accumulation factor is further calculated, and a dynamic weighting factor is generated through a nonlinear combination of shoulder activity values ​​to correct for head fatigue. Finally, combining eye fatigue and head fatigue, the degree of fatigue in a dynamic environment is calculated. This method captures the cumulative impact of task switching on user fatigue by comprehensively analyzing the frequency, complexity differences, and reorientation time of task switching, and dynamically adjusts the weight of fatigue assessment. For example, in scenarios where users frequently switch between high-complexity and low-complexity tasks, it can more realistically reflect the increase in cognitive load and avoid misjudgments caused by a single parameter. The above steps improve the scientificity and accuracy of the smart desk lamp's management of user fatigue in multi-task scenarios.

[0137] The following describes an exemplary smart desk lamp 300 provided in an embodiment of this application. Figure 3 This is an exemplary hardware structure diagram of the smart desk lamp 300 provided in this application embodiment.

[0138] In some embodiments, the smart desk lamp 300 includes a computer device. The computer device includes a processor, memory, and a network interface connected via a system bus. The processor provides computing and control capabilities. The memory includes a non-volatile storage medium and internal memory. The non-volatile storage medium stores an operating system, computer programs, and a database. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage medium. The database stores data. The network interface communicates with other external terminals or servers via a network connection. In some embodiments, the network interface can be a wired network interface; in some embodiments, it can also be a wireless network interface. When the computer program is executed by the processor, it implements the methods described in the embodiments of this application.

[0139] Those skilled in the art will understand that Figure 3 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0140] The above-described embodiments merely serve to illustrate the technical solutions of the present application, rather than limit the present application; although the present application has been described in detail with reference to the foregoing embodiments, those ordinarily skilled in the art should understand that they can still modify the technical solutions recorded in the foregoing embodiments, or equivalently replace some of the technical features thereof; and these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present application.

[0141] In the above-described embodiments, the term "when" can be interpreted as "if" or "after" or "in response to determining" or "in response to detecting" according to the context. Similarly, the phrase "upon determining" or "if detecting (the stated condition or event)" can be interpreted as "if determining" or "in response to determining" or "upon detecting (the stated condition or event)" or "in response to detecting (the stated condition or event)" according to the context.

[0142] In the above-described embodiments, all or part of the processes can be implemented by software, hardware, firmware, or any combination thereof. When implemented by software, all or part of the processes can be implemented in the form of a computer program product. The computer program product includes one or more computer instructions. When loaded and executed by a computer, the computer instructions generate all or part of the processes described in the embodiments of the present application. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another computer-readable storage medium, for example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center through a wired (such as coaxial cable, optical fiber, digital subscriber line) or wireless (such as infrared, wireless, microwave, etc.) manner. The computer-readable storage medium can be any available medium accessible by a computer or a data storage device such as a server, data center, etc. that includes one or more available media sets. The available media can be a magnetic medium (such as a floppy disk, a hard disk, a magnetic tape), an optical medium (such as a DVD), or a semiconductor medium (such as a solid-state disk), etc.

[0143] Those of ordinary skill in the art can understand that all or part of the processes in the above-described embodiments can be implemented by a computer program instructing relevant hardware, which can be stored in a computer-readable storage medium. When executed, the program can include the processes of the above-described embodiments. The aforementioned storage medium includes ROM or random access memory (RAM), a magnetic disk or an optical disk, and various media that can store program codes.

Claims

1. A smart emotional interaction method for a smart desk lamp, characterized in that, Applications in smart desk lamps include: Select the corresponding preset fatigue threshold according to the user's input instructions; obtain the user's eye features, head features, and shoulder features; If the rate of change of the head features is greater than a preset head movement threshold, determine whether the ratio of the rate of change of the shoulder features to the rate of change of the head features is within a preset head-shoulder ratio range. If not, then based on the shoulder activity level of the shoulder movement, a dynamic weighting factor that is inversely proportional to the shoulder activity level is generated, and the eye fatigue level and head fatigue level calculated by combining the eye features and head features with the dynamic weighting factor are used to calculate the dynamic environmental fatigue level. When the dynamic environment fatigue level exceeds the fatigue threshold, a voice prompt will be played. If the rate of change of the head features is not greater than the preset head movement threshold, then the gaze duration of the eyes fixed on the fixed target and the fixed posture duration of the head movement are obtained, and the corresponding gaze weight factor and fixed posture weight factor are calculated respectively. Static environment fatigue is calculated based on the eye fatigue, head fatigue, gaze weighting factor, and fixed posture weighting factor. When the static environmental fatigue level exceeds the fatigue threshold, the light brightness is adjusted; the light brightness is proportional to the static environmental fatigue level.

2. The method according to claim 1, characterized in that, The process of generating a dynamic weighting factor inversely proportional to the shoulder activity level based on the shoulder movement, and then calculating the dynamic environmental fatigue level by combining the eye fatigue level and head fatigue level calculated from the eye and head features with the dynamic weighting factor, specifically includes: When the instruction contains multiple tasks, the task types of the multiple tasks are obtained, and the task base corresponding to each task type is identified according to the preset task difference quantification rule; Monitor the user's switching operations between different tasks, record the switching frequency each time the task is switched, and the base number of tasks before and after the switch, and calculate the difference between the base number of tasks before and after the switch to obtain the degree of switching difference; The switching frequency and the degree of difference between multiple switching operations are calculated according to a preset fatigue accumulation factor calculation rule to obtain the fatigue accumulation factor; The shoulder features are quantified into shoulder movement values; The shoulder activity value and the fatigue accumulation factor are processed in a nonlinear combination manner to generate a dynamic weighting factor for head fatigue. The dynamic environmental fatigue level is calculated by multiplying the head fatigue level by the dynamic weighting factor and adding the eye fatigue level.

3. The method according to claim 2, characterized in that, The step of calculating the fatigue accumulation factor by comparing the switching frequency with multiple switching differences according to a preset fatigue accumulation factor calculation rule specifically includes: When a task switch is detected, the reorientation time of the user from the moment the first task switch is completed to the moment the user's eyes are focused on the area of ​​the second task for more than a preset gaze threshold is monitored. The task switching coefficient is obtained by calculating the ratio of the reorientation time to the preset task start time. The switching frequency and multiple switching difference degrees are calculated according to a preset fatigue accumulation factor calculation rule, and the result is multiplied by the task switching coefficient to obtain the fatigue accumulation factor.

4. The method according to claim 1, characterized in that, The process of acquiring the gaze duration of the eyes fixed on a fixed target and the fixed posture duration of the head movements, and calculating the corresponding gaze weight factor and fixed posture weight factor respectively, specifically includes: The eye fatigue factor is calculated by obtaining the duration of the user's gaze at a fixed target and the frequency of gaze deviation in the gaze direction, combined with the blink frequency and a preset adjustment factor calculation method. The head tilt angle of the head movement and the head tilt change frequency of the head tilt angle are obtained, and the head adjustment factor of the head fatigue is calculated in combination with the adjustment factor calculation method. Based on the ratio of the eye adjustment factor to the head adjustment factor, the gaze duration weight and the fixed posture duration weight of the head movements in a static environment are adjusted to obtain the gaze weight factor and the fixed posture weight factor.

5. The method according to claim 1, characterized in that, The calculation of static environment fatigue based on the eye fatigue, head fatigue, gaze weighting factor, and fixed posture weighting factor specifically includes: The gaze duration coefficient and head posture coefficient are calculated based on the difference between the gaze duration and the preset gaze duration threshold, and the difference between the head tilt angle and the preset head tilt angle threshold, respectively. Static environment fatigue is calculated based on the eye fatigue, head fatigue, gaze weighting factor, fixed posture weighting factor, fixation time coefficient, and head posture coefficient.

6. The method according to claim 1, characterized in that, When the dynamic environment fatigue level exceeds the fatigue threshold, a voice prompt is played, specifically including: When the dynamic environment fatigue level exceeds the fatigue threshold, the user's operation interval is monitored; the operation interval is when the pause time of blinking frequency exceeds a preset blink pause time threshold or the lack of change in head movement exceeds a preset head pause time threshold. If, within a preset short delay time window, the user is detected to have an operation interval that lasts longer than the preset minimum interval, a voice prompt will be played when the operation interval ends. If the operation interruption point is not detected within the short delay time window, and the dynamic environment fatigue continues to exceed the fatigue threshold, then a preset buffered voice prompt will be played after the short delay time window ends, followed by the voice prompt.

7. The method according to claim 1, characterized in that, The calculation of static environment fatigue based on the eye fatigue, head fatigue, gaze weighting factor, and fixed posture weighting factor specifically includes: When it is determined that the task type of the user's current task belongs to a preset high-precision vision task, the real-time color rendering index and real-time color temperature value of the current illumination output are obtained and compared with the preset ideal color rendering index range and color temperature stability standard to calculate the illumination quality matching factor; the illumination quality matching factor is proportional to the correlation between the real-time color rendering index and real-time color temperature value and the ideal color rendering index range and the color temperature stability standard. The eye fatigue level is corrected by combining the light quality matching factor to obtain the corrected eye fatigue level; Static environment fatigue is calculated based on the corrected eye fatigue, head fatigue, gaze weighting factor, and fixed posture weighting factor.

8. A smart desk lamp, characterized in that, The smart desk lamp includes: one or more processors and a memory; the memory is coupled to the one or more processors, the memory is used to store computer program code, the computer program code including computer instructions, and the one or more processors call the computer instructions to cause the smart desk lamp to perform the method as described in any one of claims 1-7.

9. A computer program product containing instructions, characterized in that, When the computer program product is run on the smart desk lamp, the smart desk lamp performs the method as described in any one of claims 1-7.

10. A computer-readable storage medium comprising instructions, characterized in that, When the instruction is executed on the smart desk lamp, the smart desk lamp performs the method as described in any one of claims 1-7.

Images