Biological signal driven VR scene adaptive adjustment method
Through the VR scene adaptive adjustment method driven by biosignal collection and machine learning, the problem that existing VR scenes cannot be dynamically adjusted is solved, and multi-dimensional user experience improvement and system intelligence are achieved.
Patent Information
- Application Number
- CN202510906606.X
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-07-02
- Publication Date
- 2025-10-17
AI Technical Summary
Existing VR scenes cannot be dynamically adjusted according to the user's real-time physiological and psychological state, resulting in large differences in user experience and insufficient immersion. The fixed adjustment strategy is difficult to adapt to individual differences and scene requirements, and ignores the adjustment of sensory dimensions such as touch.
By acquiring HRV data and facial micro-expressions through the bio-signal acquisition module and combining it with machine learning algorithms and feedback devices, VR scene parameters and physical stimulation can be dynamically adjusted to form a multi-dimensional adaptive adjustment system.
It achieves precise adjustment based on the user's physiological and emotional state, improves the user's immersion and experience, enhances the intelligence and stability of the system, and meets personalized needs.
Smart Images
Figure CN120808392A_ABST
Abstract
Description
TECHNICAL FIELD
[0001] The present application relates to the technical field of virtual reality, in particular to a biological signal driven VR scene adaptive adjustment method. BACKGROUND
[0002] At present, the virtual reality technology is developing rapidly, and the demand for immersive experience of users is increasing. Most of the existing VR scenes use fixed parameter settings, which cannot be dynamically adjusted according to the real-time physiological and psychological state of the user, resulting in a large difference in user experience in different physical and emotional states, and the sense of immersion is greatly discounted. For example, when the user is in different physiological states such as tension or relaxation, the fixed scene parameters cannot meet the individual needs of the user, affecting the user's sense of immersion in the virtual environment.
[0003] From the field of biological signal monitoring, although the heart rate variability coefficient obtained by the PPG sensor and the facial micro-expression recognition technology are relatively mature, these technologies are mostly applied independently and not deeply integrated with VR scene adjustment. In the prior art, the use of HRV data and facial micro-expression information is limited, and the advantages of reflecting the physiological and emotional state of the user in real time cannot be fully utilized to optimize the VR experience.
[0004] In addition, in terms of the intelligent degree of scene adjustment, the existing VR system lacks a dynamic updating mechanism based on big data and machine learning, and the adjustment strategy is fixed, which is difficult to adapt to the individual differences of different users and the changing scene requirements. Moreover, the current VR scene adjustment is mostly concentrated in the visual and auditory dimensions, and the adjustment of other sensory dimensions such as touch is less involved, the adjustment method is not comprehensive enough, and a complete closed-loop adjustment system cannot be formed, thereby limiting the sense of presence and comprehensive experience effect that users can obtain in the virtual reality environment.
[0005] Therefore, the technical personnel in the field propose a biological signal driven VR scene adaptive adjustment method to solve the above problems. SUMMARY
[0006] In view of the deficiencies of the prior art, the present application provides a biological signal driven VR scene adaptive adjustment method, which solves the problems raised in the background art.
[0007] To achieve the above purpose, the present application realizes the following technical scheme: a biological signal driven VR scene adaptive adjustment method, comprising the following steps:
[0008] The heart rate variability coefficient data of the user is obtained by the biological signal acquisition module at a preset sampling rate, and the biological signal acquisition module includes a PPG sensor;
[0009] The facial micro-expression of the user is recognized by using a facial expression recognition module, and the facial expression recognition module adopts a quantization model.
[0010] According to the obtained HRV data and the recognized facial micro-expression, a corresponding scene parameter adjustment mode and feedback device action are selected from a preset adjustment strategy library.
[0011] The VR scene is adjusted according to the selected scene parameter adjustment mode, and the feedback device is controlled to perform a corresponding action, so as to realize adaptive adjustment of the VR scene.
[0012] The method further includes a data preprocessing step, which filters, normalizes and extracts features of the obtained HRV data and facial micro-expression recognition results, so as to improve the accuracy and effectiveness of the data.
[0013] The adjustment strategy library covers multiple physiological states, each physiological state corresponds to at least one scene parameter adjustment mode and at least one feedback device action, the scene parameter adjustment mode involves adjustment of visual parameters and auditory parameters of the VR scene, and the feedback device action involves control of physical perception stimulation mode of the user.
[0014] The method establishes a user biological signal feature database, stores HRV data, facial micro-expression features and corresponding scene feedback effect records of different users in multiple VR scenes, and is used for optimizing the adjustment strategy library and improving the adjustment precision.
[0015] A machine learning algorithm is used to periodically update and optimize the adjustment strategy library, and the mapping relationship between the scene parameter adjustment mode and the feedback device action is automatically adjusted according to the continuously accumulated user biological signals and scene feedback data.
[0016] The PPG sensor is worn on a specific part of the user to detect the heart rate variation of the user and calculate the HRV data, and the input of the quantization model is the user facial image data obtained by the image acquisition device, and the output is a vector containing the user facial micro-expression category and its corresponding probability value.
[0017] Preferably, the specific sampling rate can ensure the accuracy and real-time performance of the data, and the quantization model can effectively reduce the occupation of computing resources and ensure the recognition efficiency and accuracy.
[0018] Preferably, the VR scene parameters further include one or more of brightness, volume, sound effect type and scene switching speed, and the adjustment mode is dynamically determined according to the biological signal state of the user and a preset rule, and the preset rule can be adjusted according to the feedback of different types of users and the actual use effect, so as to adapt to the needs and preferences of different users.
[0019] Preferably, the data preprocessing method can be selected and optimized according to different data characteristics and application scenarios.
[0020] Preferably, the feedback device includes a vest vibration device, a semiconductor refrigeration sheet, and other devices that can produce physical perception stimulation to the user, and the action parameters can be dynamically adjusted according to the user's biological signals. The dynamic adjustment method is based on the analysis of the user's physiological response and behavior feedback to achieve personalized feedback effect.
[0021] Preferably, the user biological signal feature database can be regularly updated and analyzed, and the machine learning algorithm can be selected and customized according to the data characteristics and optimization goals.
[0022] Preferably, the integration method of the sensor and the data fusion method can be designed and optimized according to the specific application scenario and demand to achieve the best adjustment effect.
[0023] Preferably, the method further comprises the step of real-time monitoring and evaluating the adjustment effect, and the adjustment strategy is further optimized according to the evaluation result to ensure the accuracy and effectiveness of the VR scene adjustment.
[0024] The present application provides a biological signal driven VR scene adaptive adjustment method. It has the following beneficial effects:
[0025] 1、The present application acquires the heart rate variability coefficient data and facial micro-expression information of the user through the PPG sensor and the quantization model respectively, and combines the data preprocessing step to accurately and real-timely perceive the physiological and emotional state of the user, thereby providing a basis for the precise adjustment of the VR scene. Compared with the traditional fixed VR scene setting, this method can better meet the individual needs of different users in different states, significantly improve the user's immersion and experience effect, and make the user obtain a more natural, comfortable and self-state conforming interactive feeling in the virtual environment.
[0026] 2、The present application uses the adjustment strategy library to cover a variety of physiological states, and uses the machine learning algorithm to regularly update and optimize the accumulated user data, to realize the dynamic adjustment of the scene parameter adjustment method and the action mapping relationship of the feedback device. This not only improves the adaptability and effectiveness of the adjustment strategy, and can make precise response to the individual differences of different users and the changing scene demand, but also enhances the intelligent level of the system, reduces the manual intervention, ensures the long-term stability and effectiveness of the VR scene adjustment, and provides a VR scene adaptive adjustment scheme that can continuously optimize and meet the user's needs.
[0027] 3、The biological signal acquisition is combined with VR scene control, the action of feedback device is used for assisting adjustment, such as back vibration device, semiconductor refrigeration piece and the like are used for physical perception stimulation to the user, the VR scene is comprehensively adjusted from visual, auditory, tactile and the like multidimension, a complete and closed loop adjustment system is formed. This kind of multimodal adjustment mode can more comprehensively affect the perception and emotion of user, further enhances the sense of presence and the sense of substitution of user in virtual reality environment, simultaneously, the cooperative work between each component also improves the overall performance and reliability of system, provides an innovative and practical technical solution for the field of biological signal driven VR scene adaptive adjustment. BRIEF DESCRIPTION OF DRAWINGS
[0028] Figure 1 The data processing flowchart of the application is shown in the figure.
[0029] Figure 2 The adjustment feedback flowchart of the application is shown in the figure. DETAILED DESCRIPTION
[0030] The technical solutions in the embodiments of the application will be described clearly and completely in combination with the drawings in the application specification. Obviously, the described embodiments are only part of the embodiments of the application, not all the embodiments. Based on the embodiments in the application, all other embodiments obtained by those skilled in the art without creative labor are within the protection scope of the application.
[0031] Embodiment 1:
[0032] Please refer to the attached Figure 1 -attached Figure 2 The embodiment of the application provides a biological signal driven VR scene adaptive adjustment method, which comprises the following steps:
[0033] The heart rate variation coefficient data of the user is acquired by the biological signal acquisition module at a preset sampling rate, and the biological signal acquisition module comprises a PPG sensor; the specific sampling rate can ensure the accuracy and real-time performance of the data, and the quantitative model can effectively reduce the occupation of computing resources and ensure the recognition efficiency and accuracy.
[0034] Specifically, the biosignal acquisition module is a key component for obtaining the user's heart rate variability (HRV) data, which contains a PPG sensor. The PPG sensor reflects the heart rate situation by detecting the change of blood volume, and then calculates the HRV data. The preset sampling rate refers to the frequency of data acquisition per unit time by the sensor. A higher sampling rate can capture the subtle changes in heart rate more meticulously, ensuring that the obtained HRV data is accurate and can reflect the user's current physiological state in real time. The quantization model is used to recognize facial micro-expressions. After optimization, it can effectively reduce the occupation of computing resources while ensuring the recognition efficiency and accuracy, which not only ensures that the system can quickly and accurately identify the user's micro-expression information, but also does not place too much burden on the device's computing power, ensuring the efficient operation and timeliness of data processing of the entire biosignal-driven VR scene adaptive adjustment method.
[0035] The facial expression recognition module is used to recognize the user's facial micro-expressions. The facial expression recognition module uses a quantization model.
[0036] According to the obtained HRV data and the recognized facial micro-expressions, a corresponding scene parameter adjustment method and feedback device action are selected from a preset adjustment strategy library.
[0037] The VR scene is adjusted according to the selected scene parameter adjustment method, and the feedback device is controlled to perform the corresponding action to realize the adaptive adjustment of the VR scene. The VR scene parameters also include one or more of brightness, volume, sound effect type, and scene switching speed. Their adjustment methods are dynamically determined according to the user's biosignal state and preset rules. The preset rules can be adjusted according to the feedback of different types of users and actual use effects to adapt to the needs and preferences of different users.
[0038] Specifically, the adjustment method is not fixed, but is dynamically determined according to the user's biosignal state (such as the user's physiological and psychological state reflected by the heart rate variability coefficient data and facial micro-expressions). At the same time, the preset rules can be adjusted accordingly according to the feedback of different types of users and actual use effects, so as to meet the individual needs and preferences of different users and enable each user to obtain a VR scene experience that better meets their needs. Moreover, the adjustment process is not limited to adjusting the VR scene parameters, but also controls the feedback device to perform the corresponding action, and realizes the adaptive adjustment of the VR scene through multi-dimensional adjustment, so as to improve the user's immersion and overall experience effect.
[0039] The method also includes a data preprocessing step, which filters, normalizes, and extracts features from the obtained HRV data and facial micro-expression recognition results, to improve the accuracy and effectiveness of the data. The data preprocessing method can be selected and optimized according to different data characteristics and application scenarios.
[0040] Specifically, the present application adopts a filtering data preprocessing method. Through filtering processing, useful information in the data can be retained, and these interference noises can be removed, making the data purer and providing a more accurate basis for subsequent analysis.
[0041] The adjustment strategy library covers multiple physiological states, each corresponding to at least one scene parameter adjustment mode and at least one feedback device action. The scene parameter adjustment mode involves adjusting the visual and auditory parameters of the VR scene, and the feedback device action involves controlling the physical perception stimulation mode of the user. The feedback devices include a vest vibration device, a semiconductor cooling sheet, and other devices that can produce physical perception stimulation. Their action parameters can be dynamically adjusted according to the user's biological signals, and the dynamic adjustment is based on the analysis of the user's physiological response and behavioral feedback to achieve personalized feedback effects.
[0042] Specifically, the adjustment strategy library is a collection of multiple physiological states, each with at least one corresponding scene parameter adjustment mode and at least one feedback device action. The scene parameter adjustment mode mainly adjusts the visual and auditory parameters of the VR scene to change the user's audio-visual experience in the virtual environment. At the same time, the action of the feedback device is to enhance the user's interaction experience with the virtual scene through physical perception stimulation. These feedback devices include a vest vibration device, a semiconductor cooling sheet, and other devices that can produce physical perception stimulation. Their action parameters are not fixed but can be dynamically adjusted according to the user's biological signals. This dynamic adjustment is based on in-depth analysis of the user's physiological response and behavioral feedback. In this way, the system can achieve personalized feedback effects, allowing different users to experience VR scenes that meet their needs and preferences, thereby improving user immersion and satisfaction.
[0043] The user biological signal feature database can be regularly updated and analyzed, and the machine learning algorithm can be selected and customized according to the data characteristics and optimization goals.
[0044] The method establishes a user biological signal feature database to store HRV data, facial micro-expression features, and corresponding scene feedback effect records of different users in various VR scenes, which is used to optimize the adjustment strategy library and improve the adjustment accuracy.
[0045] The integration method and data fusion method of the sensor can be designed and optimized according to the specific application scenario and requirements to achieve the best adjustment effect.
[0046] Machine learning algorithms are used to regularly update and optimize the adjustment strategy library, automatically adjusting the mapping relationship between scene parameter adjustment modes and feedback device actions based on the continuously accumulated user biological signals and scene feedback data.
[0047] Specifically, machine learning algorithms are used in the present application to periodically update and optimize the regulation strategy library. The core goal is to automatically adjust the mapping relationship between the scene parameter regulation method and the feedback device action according to the continuously accumulated user biological signal and scene feedback data. In this way, the system can adapt to different user individual differences and changing scene requirements, continuously improve the accuracy and effectiveness of the regulation strategy, and provide personalized VR scene self-adaptive regulation scheme for users.
[0048] The present application mainly uses supervised learning algorithms, including linear regression and support vector machines, where linear regression is used to predict continuous value output, predicting appropriate scene parameter values from user biological signals. Suppose we have a set of training data (x i ,y i ), where x i represents the user's biological signal feature vector, and y i represents the corresponding scene parameter value. The goal of linear regression is to find a linear model y = w T x + b, so that the error between the predicted value and the true value is minimized. The error is measured by the mean squared error (MSE):
[0049]
[0050] where n represents the number of data points, i.e. the number of samples; i represents the i-th data point, taking values from 1 to n, y i represents the true value of the i-th data point, w represents the weight parameter vector of the model. x i represents the feature vector of the i-th data point. b represents the bias parameter (intercept term) of the model.
[0051] Optimization algorithms such as gradient descent are used to minimize MSE, thereby obtaining the optimal model parameters w and b.
[0052] Support vector machines are used for classification problems, and the decision function of support vector machines is represented as:
[0053] f(x) = sign(w T x + b)
[0054] where w and b are parameters obtained by solving the optimization problem, so that the following objective function is minimized
[0055]
[0056] while satisfying the constraint y i (w T x i + b) ≥ 1 - ξ i , where C is the penalty parameter, and ξi is a slack variable.
[0057] The PPG sensor is worn on a specific part of the user's body to detect changes in the user's heart rate and calculate HRV data. The input of the quantification model is the user's facial image data obtained by the image acquisition device, and the output is a vector containing the user's facial micro-expression categories and their corresponding probability values.
[0058] Specifically, the PPG sensor is a device that detects changes in the user's heart rate, which is worn on a specific part of the user's body, such as the wrist, fingers, or earlobe. By detecting the changes in blood volume with the heart cycle, the PPG sensor can capture the user's heart rate signal and calculate the heart rate variability coefficient data from it. HRV data reflects the balance between the user's sympathetic and parasympathetic nervous systems, which can be used to assess the user's physiological stress and emotional state.
[0059] On the other hand, the quantification model is a mathematical model based on deep learning or machine learning technology, which is used to analyze and identify specific features in image data. In this process, the image acquisition device obtains the user's facial image data, which is then input into the quantification model. The quantification model is trained to recognize facial micro-expressions and output a vector containing different facial micro-expression categories and their corresponding probability values. In this way, the system can understand the user's current emotional state, providing a basis for subsequent VR scene adaptive adjustment.
[0060] The method also includes the step of real-time monitoring and evaluating the adjustment effect, which is used to further optimize the adjustment strategy based on the evaluation results to ensure the accuracy of the VR scene adjustment.
[0061] Embodiment 2
[0062] In this embodiment, the method of the present application is applied to a VR application program for immersive natural scenery tour. The program aims to make the user feel as if he is in a quiet forest or a magnificent mountain, etc. natural environment, in order to achieve the purpose of relaxing body and mind.
[0063] Data acquisition and preprocessing: a high-precision PPG sensor is selected and worn on the user's wrist to obtain the user's heart rate variability coefficient data at a sampling rate of 128Hz. At the same time, a facial expression recognition module based on MobileNetV4 quantification model is used to real-time collect the user's facial image through the camera on the VR device, and identify the facial micro-expression.
[0064] The collected HRV data is filtered to remove high-frequency noise and low-frequency drift, and then normalized to convert it to a standard range of values. Feature extraction is performed on the facial micro-expression recognition results to extract key facial feature points and expression feature vectors to improve data accuracy and effectiveness.
[0065] A regulation strategy library covering various physiological states is constructed. For example, when the HRV data is greater than 60 ms, it is determined that the user is in a relaxed state, and the corresponding scene parameter adjustment method is to increase the saturation of the VR scene by 15%, adjust the BPM (beats per minute) of the background music to 60-80, and control the vest vibration device to vibrate slightly at a frequency not exceeding 3 Hz, giving the user a comfortable and pleasant physical perception stimulation.
[0066] When the HRV data is less than 30 ms, it is determined that the user is in a tense state, and the corresponding scene parameter adjustment method is to reduce the contrast by 20%, increase the BPM of the background music to 120-140 to increase the tension and vitality of the scene, and start the semiconductor cooling sheet to provide a cooler body sensation to the user, helping them to relieve tension.
[0067] A user biological signal feature database is established to store HRV data, facial micro-expression features and corresponding scene feedback effect records of different users in various natural scenery VR scenes. Through regular mining and analysis of data in the database, machine learning algorithms are used to update and optimize the regulation strategy library. For example, through learning a large amount of user data, it is found that for some user groups, in a specific natural scenery scene, when the HRV data is in a certain intermediate range, another scene parameter adjustment method (such as adjusting the brightness and volume appropriately) can obtain better user experience, so this new regulation strategy is added to the regulation strategy library.
[0068] Adjustment effect evaluation: The adjustment effect is evaluated through user questionnaires and behavioral observation. After using the method of the present application, the user feedback on the sense of immersion in the immersive natural scenery tour VR application has been significantly improved, with 85% of users indicating that they feel as if they are truly in a natural environment. Compared with traditional fixed scene setting VR applications, the average immersion score is increased by 40%. At the same time, by analyzing the user's behavior data during use, it is found that the user's stay time in the VR scene is extended by 30%, and the number of active interactions is increased by 25%, which shows that the method of the present application can better attract the user's attention, improve their participation and interest.
[0069] Example 3
[0070] The method of the application is applied to a VR fitness application which designs various virtual fitness scenes such as virtual gyms, outdoor running scenes, etc., aiming to provide personalized fitness experiences according to the user's physical state.
[0071] Data acquisition and preprocessing: Wear PPG sensor on the user's chest to acquire HRV data at a sampling rate of 128Hz, and integrate facial expression recognition module in the front camera of VR device, use MobileNetV4 quantization model to identify user's facial micro-expression in real time. Filter the HRV data to remove interference signals during exercise, then perform feature extraction to extract key features reflecting user's heart rate variability. Normalize the facial micro-expression recognition results to improve their accuracy and stability.
[0072] Constructing adjustment strategy library, setting corresponding VR scene parameter adjustment mode and feedback device action for different HRV data range and facial micro-expression. For example, when HRV data shows that the user is in a fatigue state after high-intensity exercise, reduce the brightness and volume of the VR scene, slow down the BPM of the background music, and control the vest vibration device to vibrate at a lower frequency to simulate massage effect and help the user relax.
[0073] When the user is in an excited state during exercise (indicated by HRV data and facial micro-expression together), increase the contrast of the scene and the stimulating nature of the sound effects, increase the volume of the encouraging virtual coach prompts in the scene, and start the semiconductor cooling plate to provide a more cool and comfortable exercise environment for the user to improve their exercise performance and enthusiasm.
[0074] Establish a user biological signal feature database to store HRV data, facial micro-expression features and scene feedback effect records of users in different VR fitness scenes. Use machine learning algorithms to regularly update and optimize the adjustment strategy library. Through the analysis of the data in the database, continuously adjust and optimize the mapping relationship between scene parameter adjustment mode and feedback device action to better meet the needs of users in different fitness scenes.
[0075] Adjustment effect evaluation: By comparing the exercise data and subjective feedback of users before and after using the method of the application, it is found that the user's exercise performance has been significantly improved. After using the method of the application, the user's exercise duration is extended by an average of 20%, and the stability of exercise intensity is improved by 35%, i.e. the user can better maintain the target exercise intensity during exercise, reducing the large fluctuations in exercise intensity caused by physical discomfort or emotional fluctuations.
[0076] Meanwhile, the user satisfaction survey of the VR fitness application shows that 90% of users think that the method of the application makes the fitness process more comfortable and enjoyable, and the user's willingness to recommend the application increases by 50% compared with traditional VR fitness applications, which shows that the method of the application effectively improves the user experience and the competitiveness of the application.
[0077] Embodiment 4
[0078] This embodiment applies the method of the application to a VR education training application, which simulates various teaching scenarios such as virtual classrooms, virtual laboratories, etc., aiming to provide personalized teaching environment support according to the physiological and psychological states of different students in the learning process.
[0079] The student wears a PPG sensor on the earlobe to obtain HRV data at a sampling rate of 128Hz, and a facial expression recognition module is embedded in the camera of the VR device to identify the student's facial micro-expression during the learning process using the MobileNetV4 quantization model. The collected HRV data is filtered and normalized to eliminate noise and interference in the signal and extract effective features reflecting the student's physiological state. The facial micro-expression recognition results are extracted and optimized to improve the accuracy and reliability of the recognition results.
[0080] A regulation strategy library is constructed, and corresponding VR scene regulation strategies are set for the HRV data and facial micro-expression of students in different learning states. For example, when the student's HRV data shows that he is in a state of concentrated attention, the brightness, volume and sound effect type of the VR scene are kept stable, and the feedback device is not controlled to perform additional physical stimulation to avoid interfering with the student's learning.
[0081] When the student's HRV data and facial micro-expression indicate that he is in a state of fatigue or distraction, the brightness and volume of the scene are appropriately increased, the sound effect type is changed to more attractive and stimulating sounds, and the vest vibration device is controlled to vibrate at an appropriate frequency to remind the student to concentrate, and the student's feedback effect is recorded through the user biological signal feature database.
[0082] The data in the user biological signal feature database is regularly mined and analyzed, and machine learning algorithms are used to update and optimize the regulation strategy library. Through the learning of a large amount of student data, it is found that for students of different ages and learning subjects, different scene parameter regulation methods can better improve the learning effect of students in specific learning scenarios. For example, in the virtual laboratory scene, for physics experiment courses, when the student's facial micro-expression shows confusion and the HRV data changes, the display speed of the experimental steps is appropriately slowed down, the relevant explanation sound effects are increased, and the feedback device is controlled to give a slight vibration prompt to guide the student to better understand and operate the experiment.
[0083] Adjustment effect evaluation: By comparing the learning scores and learning performance of students before and after using the method of the application, it is found that the students' attention concentration time is significantly prolonged. After using the method of the application, the average attention concentration time of students in VR education training application is increased from 15 minutes to 25 minutes, and the learning efficiency is increased by 40%.
[0084] The method can accurately perceive the physiological and psychological state of the user according to the real-time biological signal of the user, and provide the user with more personalized, stronger immersive and more in line with the needs of the VR experience through reasonable scene parameter adjustment and feedback device action, while continuously optimizing the adjustment strategy by using the machine learning algorithm, ensuring the long-term stability and high efficiency of the system, and having wide application prospect and important practical value in the fields of VR entertainment, education, fitness and the like.
[0085] In summary: The application obtains the heart rate variability coefficient data and facial micro-expression information of the user through the PPG sensor and the quantification model respectively, and combines the data preprocessing step, which can accurately and real-timely perceive the physiological and emotional state of the user, thereby providing the basis for the accurate adjustment of the VR scene. Compared with the traditional fixed VR scene setting, this method can better meet the individual needs of different users in different states, significantly improve the immersion and experience effect of the user, and make the user obtain more natural, comfortable and in line with the state of the self in the virtual environment. Interaction experience.
[0086] Although the embodiments of the application have been shown and described, it can be understood by those of ordinary skill in the art that various changes, modifications, replacements and variations can be made to the embodiments without departing from the principles and spirits of the application, and the scope of the application is defined by the appended claims and their equivalents.
Claims
1. A bio-signal driven VR scene adaptive adjustment method, characterized in that: The following steps are involved: Acquiring the user's heart rate variation coefficient data at a preset sampling rate through a biosignal acquisition module, wherein the biosignal acquisition module includes a PPG sensor; Recognizing the user's facial micro-expressions using a facial expression recognition module, wherein the facial expression recognition module adopts a quantitative model; Based on the acquired HRV data and recognized facial micro-expressions, the corresponding scene parameter adjustment method and feedback device action are selected from the preset adjustment strategy library; Adjust the VR scene according to the selected scene parameter adjustment method, and control the feedback device to perform corresponding actions to achieve adaptive adjustment of the VR scene; The method also includes a data preprocessing step, which performs operations such as filtering, normalization, and feature extraction on the acquired HRV data and facial micro-expression recognition results to improve the accuracy and effectiveness of the data; The adjustment strategy library covers multiple physiological states, each of which corresponds to at least one scene parameter adjustment method and at least one feedback device action. The scene parameter adjustment method involves adjusting the visual parameters and auditory parameters of the VR scene, and the feedback device action involves controlling the user's physical perception stimulation method. The method establishes a user biosignal feature database to store HRV data, facial micro-expression features, and corresponding scene feedback effect records of different users in various VR scenarios, which is used to optimize the adjustment strategy library and improve adjustment accuracy. Use machine learning algorithms to regularly update and optimize the adjustment strategy library, and automatically adjust the mapping relationship between scene parameter adjustment methods and feedback device actions based on the continuously accumulated user biosignals and scene feedback data; The PPG sensor is worn on a specific part of the user to detect changes in the user's heart rate and calculate HRV data. The input of the quantization model is the user's facial image data acquired by the image acquisition device, and the output is a vector containing the user's facial micro-expression categories and their corresponding probability values.
2. The bio-signal driven VR scene adaptive adjustment method according to claim 1, characterized in that: The specific sampling rate can ensure the accuracy and real-time performance of the data, and the quantization model can effectively reduce the occupation of computing resources and ensure recognition efficiency and accuracy.
3. The bio-signal driven VR scene adaptive adjustment method according to claim 1, characterized in that: The VR scene parameters also include one or more of brightness, volume, sound effect type, and scene switching speed. The adjustment method is dynamically determined based on the user's biosignal status and preset rules. The preset rules can be adjusted based on feedback from different types of users and actual usage effects to adapt to the needs and preferences of different users.
4. The bio-signal driven VR scene adaptive adjustment method according to claim 1, characterized in that: The data preprocessing method can be selected and optimized according to different data characteristics and application scenarios.
5. The bio-signal driven VR scene adaptive adjustment method according to claim 1, characterized in that: The feedback device includes a vest vibration device, a semiconductor cooling plate and other devices that can produce physical sensory stimulation to the user. Its action parameters can be dynamically adjusted according to the user's biological signals. The dynamic adjustment method is based on the analysis of the user's physiological reactions and behavioral feedback to achieve personalized feedback effects.
6. The bio-signal driven VR scene adaptive adjustment method according to claim 1, characterized in that: The user biosignal feature database can be regularly updated and mined for analysis, and the machine learning algorithm can be selected and customized based on data characteristics and optimization goals.
7. The bio-signal driven VR scene adaptive adjustment method according to claim 1, characterized in that: The sensor integration method and data fusion method can be designed and optimized according to specific application scenarios and requirements to achieve the best adjustment effect.
8. The bio-signal driven VR scene adaptive adjustment method according to claim 1, characterized in that: The method also includes the steps of real-time monitoring and evaluation of the adjustment effect, which is used to further optimize the adjustment strategy based on the evaluation results to ensure the accuracy of VR scene adjustment.