VR simulation confrontation attention evaluation system integrated with eye movement tracking

Abstract

The invention discloses a VR simulation confrontation attention evaluation system integrated with eye movement tracking. The VR simulation confrontation attention evaluation system comprises a VR simulation training module, an eye movement tracking module, an anti-interference generation module, an attention evaluation module, a real-time feedback module, an evaluation database module and a group training management module. According to the invention, a high-precision eye movement tracking technology is deeply integrated into a VR mechanical simulation training platform with mature functions, and programmable antagonistic interference is dynamically introduced into a task, so that objective and refined evaluation of a distribution mode, stability and anti-interference capability of attention resources of a user in a complex and lifelike task environment is realized; the system overcomes the defects that a traditional evaluation method is high in subjectivity and low in ecological efficiency, quantitative attention indexes can be provided, and the user can be helped to recognize and improve the attention mode of the user through real-time feedback.

Description

Technical Field

[0001] This invention relates to the field of computer system technology, and more specifically to a VR simulated adversarial attention assessment system integrating eye tracking. Background Technology

[0002] Currently, assessments of attention states largely rely on subjective questionnaires or simple behavioral tasks, lacking objective and continuous quantitative indicators. This is particularly true for accurately capturing attention allocation and stability under complex environments or high-pressure tasks. While existing VR simulation platforms possess functions such as operational training and fault simulation, attention assessment remains at the level of basic operational recording. They fail to deeply integrate the user's visual attention characteristics with real-time environmental interference, thus hindering a systematic assessment of the scientific allocation of attention resources, resistance to interference, and task focus. Furthermore, traditional eye trackers are often used independently of VR training systems, leading to difficulties in data fusion and hindering the achievement of immediate attention feedback and interference response in immersive tasks. Summary of the Invention

[0003] To address this, the present invention provides a VR simulated adversarial attention assessment system with integrated eye tracking, in order to solve the problems of data fusion difficulties in the prior art, which make it difficult to achieve real-time attention feedback and interference response in immersive tasks.

[0004] To achieve the above objectives, the present invention provides the following technical solution:

[0005] A VR-simulated adversarial attention assessment system integrating eye tracking includes:

[0006] The VR simulation training module is used to construct a virtual reality environment containing at least one interactive task scenario. The task scenario is created based on a mechanical simulation platform and supports users to conduct immersive operation training through VR devices. The operation training includes at least one of mechanical structure cognition, principle learning, fault diagnosis, disassembly and repair, and multi-person collaborative tasks.

[0007] An eye-tracking module, integrated into the VR headset, is used to collect eye movement data of the user in real time during the VR simulation training module's task execution. The eye movement data includes gaze point coordinates, gaze duration, saccade path, pupil diameter changes, and blink frequency.

[0008] The anti-interference generation module is communicatively connected to the VR simulation training module. It is used to dynamically generate and inject visual, auditory, or task flow-related anti-interference events according to preset interference strategies during the user's task execution. The interference strategies include randomly popping up interfering objects, simulating sudden equipment failures, generating misleading operation prompts, or inserting irrelevant sound effects.

[0009] The attention assessment module is communicatively connected to the eye-tracking module and the adversarial interference generation module, respectively, and is used to calculate the user's attention assessment index based on the eye-tracking data and the user's response behavior to the adversarial interference event. The attention assessment index includes attention concentration index, visual search efficiency, interference resistance coefficient and task switching cost.

[0010] The real-time feedback module is connected in communication with the attention assessment module and is used to present the attention assessment indicators, targeted improvement suggestions and historical performance comparisons to the user in the form of visual, auditory or tactile means during or after training.

[0011] The evaluation database module stores users' historical eye-tracking data, interference event records, attention evaluation metrics, operation behavior logs, and training scenario metadata, and provides data query and model training support for the attention evaluation module.

[0012] Preferably, the VR simulation training module further includes a scene editor, which is based on a graphical zero-code editing interface of the mechanical simulation platform, allowing users to import 3D models, set task steps, define interaction logic, configure physical properties, and arrange multi-person roles and collaborative processes.

[0013] Preferably, the anti-interference generation module includes an interference strategy library and a dynamic trigger. The interference strategy library pre-stores various types of interference templates and their parameter ranges. The dynamic trigger selects a matching interference template from the interference strategy library and injects it after adjusting its intensity and timing based on the current state of the task scenario, the user's operation progress, and the real-time calculated attention load.

[0014] Preferably, the interference template includes: visual occlusion interference, used to generate semi-transparent or blurred occlusions within the critical operation field of view; cognitive conflict interference, used to provide text or voice prompts that contradict the correct operation procedure; and environmental stress interference, used to simulate time limits, virtual character urging, or increased background noise.

[0015] Preferably, the attention assessment module performs the following processing flow: First, the raw eye-tracking data is filtered and denoised, and divided into data segments corresponding to the task stage; second, the time-domain and frequency-domain features of each data segment are extracted; next, the user's behavioral response data to interference events is fused, including reaction delay, operation error type, and corrective action; then, the fused feature vector is input into the pre-trained attention assessment model, and the attention assessment index is output; finally, a comprehensive attention assessment report is generated according to the task type and assessment criteria.

[0016] Preferably, the attention evaluation model is a model trained based on machine learning algorithms, and its training data comes from multi-user historical data stored in the evaluation database module. The model can adaptively adjust the evaluation weights according to different task scenarios and supports continuous optimization through incremental learning.

[0017] Preferably, the feedback form of the real-time feedback module includes: displaying the current attention concentration in real time at the edge of the VR field of view in the form of a color spectrum or progress bar; displaying detailed evaluation charts and comparative analysis in the form of a virtual panel after the task is interrupted or ended; prompting attention distraction events through spatialized audio or controller vibration; and generating an exportable electronic report containing training suggestions.

[0018] Preferably, it also includes a group training management module, which is used by administrators to create training courses, assign training tasks, organize multi-person collaborative combat training, set interference parameters and evaluation standards, and can monitor the attention status and training progress of multiple trainees in real time, and perform group performance analysis and training resource scheduling.

[0019] The present invention has the following advantages: By deeply integrating high-precision eye-tracking technology into a mature VR mechanical simulation training platform and dynamically introducing programmable adversarial interference in the task, the present invention achieves an objective and refined evaluation of the user's attention resource allocation pattern, stability and anti-interference ability in a complex and realistic task environment.

[0020] This system overcomes the shortcomings of traditional assessment methods, such as high subjectivity and low ecological validity. It not only provides quantitative attention indicators, but also helps users recognize and improve their attention patterns through real-time feedback. The system supports model optimization based on historical data and generation of personalized interference strategies. It can be widely used in the selection, training and rehabilitation assessment of attention among military personnel, high-end equipment operators, students and special occupational groups, and has significant application value and promotion prospects. Attached Figure Description

[0021] To more intuitively illustrate the prior art and this application, exemplary drawings are provided below. It should be understood that the specific shapes and structures shown in the drawings should not generally be regarded as limiting conditions for implementing this application; for example, based on the technical concept disclosed in this application and the exemplary drawings, those skilled in the art are able to easily make conventional adjustments or further optimizations to the addition / reduction / classification, specific shapes, positional relationships, connection methods, size ratios, etc. of certain units (components).

[0022] Figure 1 A block diagram of a VR simulation adversarial attention assessment system with integrated eye tracking provided in an embodiment of this application. Detailed Implementation

[0023] The following specific embodiments illustrate the implementation of the present invention. Those skilled in the art can easily understand other advantages and effects of the present invention from the content disclosed in this specification. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. It should be understood that these embodiments are merely for further explanation of the present invention and should not be construed as limiting the scope of protection of the present invention. Technical engineers in the field can make some non-essential improvements and adjustments to the present invention based on the above-described content. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0024] Please see Figure 1 A VR-simulated adversarial attention assessment system integrating eye tracking, including:

[0025] The VR simulation training module is used to construct a virtual reality environment containing at least one interactive task scenario. The task scenario is created based on a mechanical simulation platform and supports users to conduct immersive operation training through VR devices. The operation training includes at least one of mechanical structure cognition, principle learning, fault diagnosis, disassembly and repair, and multi-person collaborative tasks.

[0026] An eye-tracking module, integrated into the VR headset, is used to collect eye movement data of the user in real time during the VR simulation training module's task execution. The eye movement data includes gaze point coordinates, gaze duration, saccade path, pupil diameter changes, and blink frequency.

[0027] The anti-interference generation module is communicatively connected to the VR simulation training module. It is used to dynamically generate and inject visual, auditory, or task flow-related anti-interference events according to preset interference strategies during the user's task execution. The interference strategies include randomly popping up interfering objects, simulating sudden equipment failures, generating misleading operation prompts, or inserting irrelevant sound effects.

[0028] An attention assessment module is connected to the eye-tracking module and the adversarial interference generation module, respectively, and is used to calculate the user's attention assessment index based on the eye-tracking data and the user's response behavior to the adversarial interference event. The attention assessment index includes attention concentration index, visual search efficiency, interference resistance coefficient and task switching cost.

[0029] The real-time feedback module, connected to the attention assessment module, is used to present the attention assessment indicators, targeted improvement suggestions, and historical performance comparisons to the user in a visual, auditory, or tactile manner during or after training.

[0030] The evaluation database module stores users' historical eye-tracking data, interference event records, attention evaluation metrics, operation behavior logs, and training scenario metadata, and provides data query and model training support for the attention evaluation module.

[0031] During implementation, this system provides a highly realistic mechanical operation environment through a VR simulation training module. The eye-tracking module is directly integrated into the VR device, seamlessly capturing the user's actual visual attention allocation during tasks, providing an objective and continuous physiological data source for evaluation. The anti-interference generation module is a key innovation; it proactively and dynamically injects controllable interference into a smooth task flow, simulating unexpected events and pressures in a real work environment, thereby stimulating and testing the user's attention resilience. The attention assessment module, acting as the system's "brain," fuses and analyzes eye-tracking data and behavioral response data, transforming the abstract concept of "attention" into quantifiable indicators (such as concentration and anti-interference coefficient) for scientific evaluation. The real-time feedback module instantly visualizes the evaluation results, allowing users to immediately recognize the strengths and weaknesses of their attention patterns, forming a closed loop of "training-evaluation-feedback," significantly improving training efficiency. The evaluation database module supports long-term tracking, personalized modeling, and algorithm optimization. This system upgrades traditional operational skills training to the quantitative assessment and shaping of higher-order cognitive abilities (attention), making it particularly suitable for the training and selection of professionals who need to maintain focus under high pressure and high interference environments.

[0032] The VR simulation training module also includes a scene editor, which is based on a graphical zero-code editing interface of the mechanical simulation platform. This allows users to import 3D models, set task steps, define interaction logic, configure physical properties, and arrange multi-person roles and collaborative processes.

[0033] The scenario editor leverages its graphical, zero-code nature to significantly lower the technical barrier to building complex evaluation scenarios. Administrators or instructors, without programming knowledge, can quickly drag and drop to build training tasks containing specific machinery, specific faults, and specific collaborative processes, and flexibly bind interference trigger points, key observation areas, and task steps. This ease of use and flexibility allows for the rapid customization of customized evaluation schemes based on different evaluation objectives (such as evaluating attention to a specific instrument or evaluating attention allocation in team collaboration), broadening the system's application scope and reducing deployment costs.

[0034] The anti-interference generation module includes an interference strategy library and a dynamic trigger. The interference strategy library pre-stores various types of interference templates and their parameter ranges. The dynamic trigger selects a matching interference template from the interference strategy library and injects it after adjusting its intensity and timing based on the current state of the task scenario, the user's operation progress, and the real-time calculated attention load.

[0035] The interference strategy library templates and parameterizes various interference events for easy management and invocation. Dynamic triggers intelligently decide the type, intensity, and timing of interference based on the real-time context of the task (such as the criticality of the user's current operation and the duration of sustained focus), rather than simply triggering them randomly. This intelligent and adaptive method of interference generation provides "just the right" challenge for users of different skill levels or states, avoiding both weak interference that renders the evaluation meaningless and overly strong interference that leads to user frustration, thus achieving more accurate and personalized attention stress testing.

[0036] The interference templates in the interference strategy library include: visual occlusion interference, used to generate semi-transparent or blurred occlusions within the critical operation field of view; cognitive conflict interference, used to provide text or voice prompts that contradict the correct operation procedure; and environmental stress interference, used to simulate time limits, virtual character urging, or increased background noise.

[0037] The aforementioned interferences target different dimensions of attention resources—perceptual channels, decision-making logic, and emotional load—forming a multi-dimensional, three-dimensional system to combat interference. Its advantage lies in its ability to more comprehensively simulate the sources of complex interference in the real world, thereby systematically evaluating and training users' ability to maintain and shift attention under different types of interference. This makes the evaluation results more ecologically valid, and the training effects more transferable to actual work scenarios.

[0038] The attention assessment module performs the following processing flow: First, the raw eye-tracking data is filtered and denoised, and divided into data segments corresponding to the task stage; second, the time-domain and frequency-domain features of each data segment are extracted; next, the user's behavioral response data to interference events is fused, including reaction delay, operation error type, and corrective action; then, the fused feature vector is input into the pre-trained attention assessment model, and the attention assessment index is output; finally, a comprehensive attention assessment report is generated according to the task type and assessment criteria.

[0039] This process, by deeply integrating eye-tracking physiological signals with external behavioral data rather than relying on a single type of data, significantly improves the accuracy and robustness of the assessment. For example, a user may gaze at the correct location (good eye-tracking data) but perform an incorrect action (poor behavioral data). Fusion analysis can reveal this cognitive disconnect of "knowing and doing," which cannot be detected by analysis of a single data source, thus providing a deeper insight into cognitive states.

[0040] The attention evaluation model is a model trained based on machine learning algorithms. Its training data comes from the multi-user historical data stored in the evaluation database module. The model can adaptively adjust the evaluation weights according to different task scenarios and supports continuous optimization through incremental learning.

[0041] By training with massive amounts of historical data in the evaluation database, the model learns the complex relationships between attention performance and eye movement and behavioral characteristics in different task scenarios, thus enabling the system to continuously evolve: the longer it is used and the more data it accumulates, the higher the model's evaluation accuracy and its ability to distinguish between different users becomes. Simultaneously, adaptive weight adjustment allows the same system to fairly and effectively evaluate scenarios of varying difficulty, from simple cognitive tasks to complex team collaborations, demonstrating extremely high versatility.

[0042] The real-time feedback module provides feedback in the following forms: displaying the current level of attention in real time at the edge of the VR field of view in the form of a color spectrum or progress bar; displaying detailed evaluation charts and comparative analysis in the form of a virtual panel after a task is interrupted or completed; providing alerts for attention distraction events through spatialized audio or controller vibration; and generating exportable electronic reports containing training suggestions.

[0043] Real-time spectral display at the edge of the VR field of view allows users to perceive changes in their attention levels without cognitive load during tasks; detailed virtual panels after tasks provide in-depth review materials; and spatial audio and vibration serve as immediate alerts. This immediate, intuitive, and rich feedback mechanism greatly enhances users' metacognitive abilities (awareness of their own cognitive processes), helping them quickly establish correct attention patterns, transforming tedious assessments into an immersive learning experience, and effectively improving training motivation and effectiveness.

[0044] It also includes a group training management module, which is used by administrators to create training courses, assign training tasks, organize multi-person collaborative combat training, set interference parameters and evaluation standards, and monitor the attention status and training progress of multiple trainees in real time, and perform group performance analysis and training resource scheduling.

[0045] The group training management module allows instructors to centrally configure, monitor, and manage multi-person training, enabling large-scale, standardized attention assessment and training. Instructors can uniformly set interference schemes and evaluation criteria, while observing the attention distribution and collaborative efficiency of the entire team in collaborative tasks, identifying weaknesses in the team's attention structure. This is crucial for highly collaborative collective attention training in military, emergency, and surgical teams, enabling the optimization of the team's overall cognitive resource allocation at a systemic level.

[0046] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A VR simulated adversarial attention assessment system integrating eye tracking, characterized in that, include: The VR simulation training module is used to construct a virtual reality environment containing at least one interactive task scenario. The task scenario is created based on a mechanical simulation platform and supports users to conduct immersive operation training through VR devices. The operation training includes at least one of mechanical structure cognition, principle learning, fault diagnosis, disassembly and repair, and multi-person collaborative tasks. An eye-tracking module, integrated into the VR headset, is used to collect eye movement data of the user in real time during the VR simulation training module's task execution. The eye movement data includes gaze point coordinates, gaze duration, saccade path, pupil diameter changes, and blink frequency. The anti-interference generation module is communicatively connected to the VR simulation training module. It is used to dynamically generate and inject visual, auditory, or task flow-related anti-interference events according to preset interference strategies during the user's task execution. The interference strategies include randomly popping up interfering objects, simulating sudden equipment failures, generating misleading operation prompts, or inserting irrelevant sound effects. The attention assessment module is communicatively connected to the eye-tracking module and the adversarial interference generation module, respectively, and is used to calculate the user's attention assessment index based on the eye-tracking data and the user's response behavior to the adversarial interference event. The attention assessment index includes attention concentration index, visual search efficiency, interference resistance coefficient and task switching cost. The real-time feedback module is connected in communication with the attention assessment module and is used to present the attention assessment indicators, targeted improvement suggestions and historical performance comparisons to the user in the form of visual, auditory or tactile means during or after training. The evaluation database module stores users' historical eye-tracking data, interference event records, attention evaluation metrics, operation behavior logs, and training scenario metadata, and provides data query and model training support for the attention evaluation module.

2. The VR simulated adversarial attention assessment system integrating eye tracking according to claim 1, characterized in that, The VR simulation training module also includes a scene editor, which is based on a graphical zero-code editing interface of the mechanical simulation platform. This allows users to import 3D models, set task steps, define interaction logic, configure physical properties, and arrange multi-person roles and collaborative processes.

3. The VR simulated adversarial attention assessment system integrating eye tracking according to claim 1, characterized in that, The anti-interference generation module includes an interference strategy library and a dynamic trigger. The interference strategy library pre-stores various types of interference templates and their parameter ranges. The dynamic trigger selects a matching interference template from the interference strategy library and injects it after adjusting its intensity and timing based on the current state of the task scenario, the user's operation progress, and the real-time calculated attention load.

4. The VR simulated adversarial attention assessment system with integrated eye tracking according to claim 3, characterized in that, The interference templates include: visual occlusion interference, used to generate semi-transparent or blurred occlusions within the critical operational field of view; cognitive conflict interference, used to provide text or voice prompts that contradict the correct operating procedures; and environmental stress interference, used to simulate time limits, virtual character urging, or increased background noise.

5. The VR simulated adversarial attention assessment system integrating eye tracking according to claim 1, characterized in that, The attention assessment module performs the following processing flow: First, the raw eye-tracking data is filtered and denoised, and divided into data segments corresponding to the task stage; second, the time-domain and frequency-domain features of each data segment are extracted; next, the user's behavioral response data to interference events is fused, including reaction delay, operation error type, and corrective action; then, the fused feature vector is input into the pre-trained attention assessment model, and the attention assessment index is output; finally, a comprehensive attention assessment report is generated according to the task type and assessment criteria.

6. The VR simulated adversarial attention assessment system integrating eye tracking according to claim 5, characterized in that, The attention evaluation model is a model trained based on machine learning algorithms. Its training data comes from the multi-user historical data stored in the evaluation database module. The model can adaptively adjust the evaluation weights according to different task scenarios and supports continuous optimization through incremental learning.

7. The VR simulated adversarial attention assessment system integrating eye tracking according to claim 1, characterized in that, The real-time feedback module provides feedback in the following forms: displaying the current level of attention in real time at the edge of the VR field of view in the form of a color spectrum or progress bar; displaying detailed evaluation charts and comparative analysis in the form of a virtual panel after a task is interrupted or completed; providing alerts for attention distraction events through spatialized audio or controller vibration; and generating exportable electronic reports containing training suggestions.

8. The VR simulated adversarial attention assessment system integrating eye tracking according to claim 1, characterized in that, It also includes a group training management module, which is used by administrators to create training courses, assign training tasks, organize multi-person collaborative combat training, set interference parameters and evaluation standards, and monitor the attention status and training progress of multiple trainees in real time, and perform group performance analysis and training resource scheduling.

Images