HK1241070B - Control method of biological characteristic collection hardware and apparatus - Google Patents

Description

Control method and device for biometric feature acquisition hardware

Technical Field

The present application relates to the field of computer applications, and in particular to a method and device for controlling biometric feature acquisition hardware.

Background Art

With the continuous development of the mobile internet, it is becoming increasingly common to equip smart terminals with biometric hardware to collect user biometrics for identity authentication. Collecting user biometrics for identity authentication simplifies the authentication process and reduces the complexity of user authentication. For example, by introducing a biometric-based authentication mechanism, users can perform services requiring identity authentication through smart terminals without having to perform complex interactions such as entering a password.

In actual applications, since the biometric collection hardware installed on the smart terminal may increase the power consumption of the smart terminal, the biometric collection hardware is usually turned off normally and is only started when needed. For example, in some scenarios, if a dedicated biometric camera (such as an iris camera) is installed in the smart terminal, the biometric camera is only started when it is needed to collect the user's eye pattern features or iris features to authenticate the user's identity.

However, when starting the biometric feature acquisition hardware, there is usually a hardware initialization delay. For example, taking the biometric camera as an example, for the biometric camera module built into the smart terminal, it takes about 2 seconds (the size of the delay depends on the specific device) from the time it is turned on to the time the initialization is completed and the image is output. For the camera module external to the smart terminal via USB, it takes about 3 seconds from the time it is turned on to the time the initialization is completed and the image is output. It can be seen that in the existing scheme for authenticating user identity based on the user's biometric features, it is not possible to take into account the balance between device power consumption and the delay experience caused to the user by hardware initialization.

Summary of the Invention

This application proposes a method for controlling biometric feature collection hardware, which is applied to a client. The method includes:

Collect user operation behavior data;

Pre-determining whether a target service meets a preset trigger condition based on the collected operation behavior data; wherein the target service is a user service that requires security authentication based on the user's biometric characteristics;

If it is pre-determined that the target service meets the preset triggering condition, the preset biometric feature collection hardware is started.

This application also proposes a control device for biometric feature collection hardware, which is applied to a client, and the device includes:

Collection module, collects user operation behavior data;

A prediction module, which predicts whether a target service meets a preset trigger condition based on the collected operation behavior data; wherein the target service is a user service that requires security authentication based on the user's biometric characteristics;

The startup module starts the preset biometric feature collection hardware if it is pre-determined that the target service meets the preset triggering condition.

In this application, based on the collected user operation behavior data, it is predicted whether the target business that needs to perform security authentication based on the user's biometrics meets the trigger conditions, and when it is predicted that the target business meets the trigger conditions, the biometric collection hardware is immediately started;

On the one hand, due to the use of a prediction mechanism for the time when the target service is triggered, the biometric collection hardware can be started in advance when performing security authentication for the above target service, thus ensuring that users do not feel the hardware initialization delay of the biometric collection hardware, thereby improving the user experience;

On the other hand, since the biometric collection hardware remains turned off by default, it will only be turned on when it is predicted that the target business meets the trigger conditions. Therefore, while ensuring that users do not feel the hardware initialization delay of the biometric collection hardware, the power consumption of the device can be taken into account as much as possible.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG1 is a flow chart of a method for controlling biometric feature acquisition hardware according to an embodiment of the present application;

FIG2 is a logic block diagram of a control device for biometric feature collection hardware provided by an embodiment of the present application;

FIG3 is a hardware structure diagram of a terminal device carrying a control device for biometric feature collection hardware according to an embodiment of the present application.

DETAILED DESCRIPTION

This application aims to propose a method for introducing a prediction mechanism for whether the target service meets the triggering conditions in an application scenario of performing security authentication on the target service triggered by the user based on biometrics, and based on the prediction result, to start the biometric collection hardware carried by the smart terminal in advance, thereby optimizing the delay experience caused to the user due to the hardware initialization delay of the biometric collection hardware, and at the same time, on this basis, taking into account the power consumption of the smart terminal as much as possible. Technical solution.

For example, when the technical solution of the present application is applied to a quick payment scenario in a VR (Virtual Reality) scenario, the above-mentioned biometric feature collection hardware can specifically be a biometric camera installed in a VR terminal; when a user triggers a quick payment service in a VR scenario, the VR client can call the above-mentioned biometric camera installed in the VR terminal to collect the user's eye pattern features or iris features, and quickly complete security authentication for the payment service triggered by the user in the VR scenario.

In this scenario, in order to optimize the delay experience caused to the user by the hardware initialization delay of the biometric hardware, the VR client can predict whether the above-mentioned quick payment service meets the trigger conditions based on the user's operation behavior data. When it is predicted that the quick payment service meets the trigger conditions and is about to be triggered by the user, the above-mentioned biometric camera can be turned on in advance, thereby ensuring that the user does not feel the hardware initialization delay of the biometric camera, thereby improving the user experience; for example, when the user is purchasing goods in the VR scene, when the product purchase interface is displayed (that is, when it is predicted that the user may be about to perform a payment operation), the VR client can turn on the biometric camera in advance in the background, so that when the user confirms the payment, the biometric camera has been initialized, so that the user will not feel the delay experience caused by the process from turning on the biometric camera to completing the hardware initialization and outputting the image.

The present application is described below through specific embodiments and in combination with specific application scenarios.

Please refer to FIG1 , which shows a method for controlling biometric feature collection hardware according to an embodiment of the present application. The method is applied to a client and performs the following steps:

Step 101: Collect user operation behavior data;

Step 102: Pre-determine whether a target service meets a preset trigger condition based on the collected operation behavior data; wherein the target service is a user service that requires security authentication based on the user's biometric characteristics;

Step 103: If it is determined that the target service meets the preset triggering condition, the preset biometric feature collection hardware is started.

The target service may include any type of user service that requires security authentication based on the user's biometrics; for example, in actual applications, the target service may be an online payment service initiated by the user through the client.

The above-mentioned client may include client software installed on the user's smart terminal, which can provide services related to the above-mentioned target business to the user; for example, in one application scenario, the above-mentioned client may be a VR client, and the above-mentioned target business may be a quick payment business initiated by the user in the VR scene.

The above-mentioned biometric features may include any type of biometric features that can verify the identity of the user; corresponding to the above-mentioned biometric features, the above-mentioned biometric feature collection hardware may include related hardware for collecting the above-mentioned biometric features of the user; wherein, the biometric feature hardware may specifically be a hardware module built into the user's smart terminal, or it may be a hardware module external to the user's smart terminal through methods such as USB.

For example, in actual applications, the above-mentioned biometric features may include the user's iris features or eye pattern features, and the above-mentioned biometric feature collection hardware may specifically be a built-in or external biometric recognition camera in the user's smart terminal.

The aforementioned operation behavior data may specifically be the operation behavior performed by the user to trigger the aforementioned target service, and may be used to predict whether the aforementioned target service is about to be triggered;

For example, taking the quick payment service scenario in a VR scene as an example, users can usually trigger quick payment in the VR scene by issuing voice commands, gazing at the user options for triggering the quick payment service provided in the VR scene with visual focus, or passing through the area where the user options for triggering the quick payment service are located by moving the visual focus; therefore, in this scenario, the above-mentioned operation behavior data may specifically include voice command fragments issued by the user, or the movement trajectory data of the user's visual focus, etc.

The following describes the technical solution of this application in detail, taking the client as a VR client and combining the application of the technical solution shown in steps 101 to 103 in a VR scenario as an example. Obviously, the use of the client as a VR client is only exemplary and is not intended to specifically limit the technical solution of this application.

The following describes the technical solution of this application in detail through four stages: VR scene model creation, collection of operation behavior data, prediction of target business triggering, and control of user biometric collection hardware.

1) VR scene model creation.

In this example, developers can use specific modeling tools to create VR scene models. The above modeling tools are not specifically limited in this example; for example, developers can use more mature modeling tools such as Unity, 3dsMax, Photoshop, etc. to create VR scene models.

Among them, when developers use modeling tools to create VR scene models, the VR scene model and the texture map of the VR scene can all be derived from real scenes in real life; for example, you can use video to collect material texture maps and plane models of real scenes in advance, and then use modeling tools such as Photoshop or 3dmax to process the texture and build a three-dimensional model of the real scene, and then import it into the unity3D platform (U3D for short), and render the picture in multiple dimensions such as sound effects, graphical interface, plug-ins, and lighting on the U3D platform, and then write interactive code to finally complete the modeling of the VR scene model.

In this example, in addition to creating a VR scene model, in order to enable users to perform the above-mentioned target business in the VR scene, the developer can also use the above-mentioned modeling tool to create a 2D or 3D business interface corresponding to the above-mentioned target business in the above-mentioned VR scene model.

For example, in one embodiment, the business interface can be a quick payment interface created using the modeling tool, such as a virtual cashier interface. Users can interact with the payment interface through specific interactions (such as focusing on the payment interface) to complete quick payments in the VR scene.

2) Collection of operational behavior data.

In a VR scenario, users can trigger the above target services in the VR scene in the following three ways:

First, users can trigger the above-mentioned target services in a VR scenario by issuing voice commands.

For example, taking the above-mentioned target business as the quick payment business in the VR scenario, the VR client can be equipped with a voice recognition module, and the user can issue a customized voice command for triggering the quick payment business while wearing the VR terminal for an immersive experience, thereby initiating quick payment in the VR scenario in a more natural interactive way.

In the case shown above, the above-mentioned operation behavior data may include voice command data issued by the user when triggering the above-mentioned target service through voice operation in the VR scene; for example, voice command fragments.

In this case, when the user interacts with the VR client by issuing voice commands, the VR client can use the relevant voice acquisition hardware to collect the voice command fragments issued by the user as calculation parameters for subsequent prediction of whether the above-mentioned target business meets the trigger conditions.

The second method is to provide an interactive option (such as an interactive button) for triggering the above-mentioned target service in the VR scene, and the user can stare at the interactive option in the VR scene with visual focus to select the interactive option, thereby triggering the above-mentioned target service.

For example, still taking the above-mentioned target business as the quick payment business in the VR scene as an example, a payment button for triggering quick payment can be provided in the VR scene. The user can control the displacement of the visual focus, move the visual focus to the area where the payment button is located, and keep staring at it, select the payment button by staring, and then initiate quick payment in the VR scene.

Thirdly, an interactive option for triggering the above-mentioned target business can also be provided in the VR scene, and the user can control the displacement of the visual focus and move the visual focus through the area where the interactive option is located to select the interactive option, thereby triggering the above-mentioned target business.

For example, still taking the above-mentioned target business as the quick payment business in the VR scene as an example, a payment button for triggering quick payment can still be provided in the VR scene. The user can control the displacement of the visual focus and control the movement trajectory of the visual focus to pass through the area where the payment button is located (the specific passing method is not specifically limited in this application, and technical personnel in this field can refer to the records in the relevant technology when implementing it) to select the payment button and then initiate quick payment in the VR scene.

In the two cases shown above, the above-mentioned operation behavior data may include the movement trajectory data of the user's visual focus when the user triggers the above-mentioned target service through visual focus operation in the VR scene.

In this scenario, as the user interacts with the VR client by controlling the movement of their visual focus, the VR client can record the coordinate data of the user's visual focus in the VR scene and the corresponding occurrence time in real time. Then, based on the corresponding occurrence time of each recorded coordinate data, all recorded historical coordinate data is organized into a time series to restore the user's visual focus movement trajectory data. This organized time series, the visual focus movement trajectory data, can be used as a calculation parameter to subsequently predict whether the target service meets the trigger conditions.

Of course, in actual applications, the above-mentioned user operation behavior data may include not only the voice instruction data and the movement trajectory data of the user's visual focus shown above, but also other types of operation behavior data.

That is, in this application, all behavioral data corresponding to the operational behavior performed by the user that can be used to trigger the above-mentioned target can be used as operational behavior data to predict whether the above-mentioned target business is about to be triggered, and will not be listed one by one in this application.

3) Prediction of target business triggering.

In this example, the above-mentioned operation behavior data may specifically include data types such as the voice command fragment issued by the user when triggering the above-mentioned target service, and the movement trajectory data of the user's visual focus.

Among them, when the VR client takes the above-mentioned operation behavior data as input and predicts whether the above-mentioned target business meets the preset trigger conditions, the specific content of the trigger conditions can correspond to the data type corresponding to the above-mentioned operation behavior data. For different data types, different trigger conditions can be configured for the above-mentioned target business.

The following describes the process of predicting whether the target service meets the preset triggering conditions, taking the above-mentioned operation behavior data as voice command segments and visual focus movement trajectory data as examples.

1. Prediction based on voice command fragments

In one embodiment shown, when the operation behavior data is a voice instruction segment, the preset trigger condition may specifically be "the similarity probability between the voice instruction segment and the complete voice instruction is greater than a preset similarity threshold.

Specifically, when a user triggers the above-mentioned target service through voice commands, the voice commands issued by the user are usually composed of several voice command fragments according to a certain semantic relationship; therefore, in order to predict in advance whether the target service meets the triggering conditions before the VR client successfully triggers the above-mentioned target service based on the complete voice command issued by the user, the VR client can be equipped with a voice analysis matching model for calculating the similarity probability of the voice command fragment corresponding to the complete voice command, input the received voice command fragment issued by the user into the model for calculation, and obtain the similarity probability of the voice command fragment corresponding to the completed voice command, and predict whether the above-mentioned target service meets the triggering conditions based on whether the calculated similarity probability is greater than the command similarity threshold.

Among them, the specific type of the above-mentioned speech analysis and matching model is not particularly limited in this application. In practical applications, it can be selected based on actual needs; for example, in one embodiment shown, the above-mentioned speech analysis and matching model can be an HMM (Hidden Markov Model) model, or other types of statistical analysis models with the same function based on the Viterbi algorithm.

When a user triggers the above-mentioned target service by issuing a voice command, the VR client can parse the voice command fragments issued by the user one by one in real time, and parse the first voice command fragment issued by the user into a corresponding character string fragment; wherein, the character string fragment is a subset of the character string parsed from the complete voice command, specifically, it can be a unit character in the character string parsed from the complete voice command, or a character fragment composed of multiple characters divided according to semantics.

After the first voice command fragment issued by the user is parsed into a corresponding string fragment, the VR client can use the string fragment as an input parameter and input it into the above-mentioned voice parsing matching model for calculation to obtain the similarity probability of the first voice command fragment corresponding to the complete command string parsed from the complete voice command, and then compare the calculated similarity probability with a preset similarity threshold; wherein, the similarity threshold is used to measure whether the calculated similarity probability can successfully trigger the above-mentioned target business. In actual application, it can be customized based on the specific prediction accuracy requirements.

If the calculated similarity probability is greater than the similarity threshold, it can be determined that the above-mentioned target service meets the trigger conditions and is about to be triggered by the user; conversely, if the calculated similarity probability is less than or equal to the instruction similarity threshold, it indicates that based on the first voice instruction segment issued by the user, it is predicted that the above-mentioned target service does not currently meet the trigger conditions.

In this case, the VR client can continue to parse the next voice command segment issued by the user into the corresponding string segment in the same manner, and then merge the string segment corresponding to the first voice command segment with the string segment corresponding to the next voice command segment issued by the user. Then, in the same manner, re-input the merged string segments into the above-mentioned voice parsing matching model for calculation to obtain the similarity probability of the merged voice command segment corresponding to the complete command string parsed from the complete voice command, and then re-compare the calculated similarity probability with the preset similarity threshold, and predict whether the above-mentioned target business meets the trigger condition based on the comparison result.

Similarly, if based on the merged voice instruction segment, it is still impossible to predict whether the above-mentioned target service is about to be triggered by the user, that is, the similarity probability of the merged voice instruction segment corresponding to the complete instruction string parsed from the complete voice instruction is still less than or equal to the above-mentioned similarity threshold, then based on the same operation, the merged voice instruction segment can be merged in sequence with the string segment parsed from the next voice instruction segment received later, and then the same prediction process is performed until the merged voice instruction segment matches the complete voice instruction segment parsed from the complete voice instruction (that is, after the complete voice instruction is merged, the above-mentioned target service has been triggered, and the prediction process has ended).

For example, taking the above-mentioned target business as the quick payment business in the VR scene as an example, assuming that the quick payment is triggered by the customized voice command "Open Sesame" in the VR scene, then in the process of the user triggering the quick payment by issuing the voice command, the VR client can first calculate the similarity probability of the first voice command segment "Zhi" issued by the user corresponding to the complete voice command "Open Sesame" according to the prediction method shown above, and then predict whether the quick payment business is about to be triggered by the user based on the similarity probability; if the prediction fails, the VR client can continue to merge the first voice command segment "Zhi" with the next voice command segment "Ma" into "Zhima", and then continue to perform similar prediction processing. If the prediction still fails based on the merged voice command segment "Zhima", the merged voice command segment "Zhima" and the next voice command segment "Kai" can be further merged into the voice command segment "Zhima Kai", and then continue to perform similar prediction processing until the merged command segment completely matches the complete command segment.

In this way, the VR client can predict whether the above-mentioned target service meets the triggering conditions in advance before the user has fully issued a complete voice command, so that the prediction result can be used to start the biometric collection hardware equipped with the VR terminal where the VR client is located in advance.

2. Trajectory Data Prediction Based on Visual Focus

In another embodiment shown, when the above-mentioned operation behavior data is the movement trajectory data of the user's visual focus, the above-mentioned preset trigger condition can specifically be "the movement trajectory of the user's visual focus from the current moment to the preset time threshold predicted based on the movement trajectory data of the user's visual focus falls into the preset area where the interactive option for triggering the above-mentioned target business is located."

In this case, when predicting whether the above-mentioned target service is about to be triggered based on the trajectory data of the visual focus, it can be achieved by predicting whether the movement trajectory of the user's visual focus from the current moment to N seconds later (N is the above-mentioned preset time threshold) enters the area where the above-mentioned interactive option is located.

During implementation, in the above-mentioned VR scene, an interactive option for triggering the above-mentioned target business can be provided; for example, if the above-mentioned target business is a quick payment business in the VR scene, then the interactive option can specifically be a "Buy Now" interactive button provided in the product interface in the VR scene.

Specifically, the VR client can be pre-loaded with a trained prediction model based on the historical movement trajectory data of the user's visual focus; the prediction model can be used to predict the movement trajectory of the user's visual focus from the current moment to N seconds later.

Among them, the specific type of the above-mentioned prediction model is not particularly limited in this application. It can be a deep learning model built based on a neural network, or it can be a prediction model built based on a specific prediction algorithm; for example, in one embodiment shown, the above-mentioned prediction model can be a prediction model built based on the Kalman filter prediction equation.

After the client successfully restores the movement trajectory data of the user's visual focus based on the recorded historical coordinates of the user's visual focus and the corresponding occurrence time, the movement trajectory data can be used as input data and input into the above-mentioned prediction model for calculation to predict the movement trajectory of the user's visual focus from the current moment to N seconds later. After the movement trajectory from the current moment to N seconds is predicted, it can be further determined whether the predicted movement trajectory falls within the area where the above-mentioned interactive options are located;

If the predicted movement trajectory falls into the area where the above-mentioned interactive option is located, the VR client can determine that the above-mentioned target service meets the above-mentioned triggering conditions, and the target service is about to be triggered by the above-mentioned user; otherwise, it can be determined that this prediction has failed.

In addition, in order to improve the accuracy of prediction and prevent user misoperation, a valid area can be set in the area where the above-mentioned interactive options are located; wherein, the valid area can be a sub-area centered in the area where the interactive options are located; for example, it can be the center 50% area of the area where the above-mentioned interactive options are located.

In this way, only when the predicted movement trajectory falls into the valid area of the area where the interactive option is located, will it be determined that the movement trajectory falls into the area where the interactive option is located, thereby effectively preventing the occurrence of user misoperation.

Of course, in actual applications, since the user's visual focus is not a very standard straight line in most cases, even if the VR client predicts the movement trajectory from the current moment to N seconds later, it may be difficult to determine whether the movement trajectory can enter the area where the above-mentioned interactive options are located.

In this case, the above-mentioned prediction model carried by the VR client can further introduce a relevant algorithm for calculating the probability that the predicted movement trajectory enters the area where the above-mentioned interactive option is located. Therefore, after the VR client predicts the above-mentioned movement trajectory, it can only further output a probability of entering the area where the above-mentioned interactive option is located through the preset model, and then it can be determined whether the movement trajectory has entered the area where the above-mentioned interactive option is located based on whether the calculated probability value reaches the threshold.

The algorithm for calculating the probability of the movement trajectory entering the area where the interactive option is located is not particularly limited in this example.

For example, in one embodiment shown, if the above-mentioned prediction model is a Kalman filter prediction model, then based on the results of the movement trajectory predicted by the model, an approximate integral calculation can be further performed to calculate the probability of the movement trajectory entering the area where the above-mentioned interactive option is located, and the calculated probability value can be output.

In addition, it should be noted that the specific value of the above N value is not particularly limited in this application and can be customized based on needs in actual applications;

For example, in one implementation shown, the initial value of the N value can be set to the delay duration for the above-mentioned biometric acquisition hardware to perform hardware initialization; for example, assuming that the above-mentioned biometric acquisition hardware performs hardware initialization, it takes a 2-second delay from startup to completion of initialization, then the initial value of the above-mentioned N value can also be set to 2 seconds.

In another illustrated embodiment, when the operational behavior data is the movement trajectory data of the user's visual focus, the preset trigger condition may specifically be "the user's visual focus, as determined based on the movement trajectory data, falls within a preset area containing an interactive option for triggering the target service, and the duration of the focus exceeds a gaze wait time threshold." In this case, when predicting whether the target service is about to be triggered based on the visual focus trajectory data, this can be achieved by calculating the duration of the user's visual focus after it falls within the area containing the interactive option, and then determining whether this duration of focus reaches the preset gaze wait time threshold.

In a VR scenario, users can usually control their visual focus to stay in the area where the above-mentioned interactive options are located, keep hovering, and select the above-mentioned interactive options by "staring", thereby triggering the above-mentioned target business.

However, in related technologies, when a user selects the above-mentioned interactive option by "gazing", the time that the user's visual focus stays in the area where the above-mentioned interactive option is located is usually composed of the gaze waiting time (denoted as T1) and the gaze confirmation time (denoted as T2).

When the user's visual focus enters the area where the above-mentioned interactive option is located, the VR client can count the length of time the visual focus stays in the area where the interactive option is located in the background. When the value of the stay time is greater than T1, the gaze waiting ends at this time, and the VR client can confirm that the user's current operation is a valid "gaze" operation in the background, and output a related waiting prompt in the area where the interactive option is located; at the same time, the VR client can re-time from this moment and count the length of time the visual focus stays in the area where the interactive option is located. If the stay time obtained after re-timing is greater than T2, the gaze confirmation ends at this time, and the VR client can select the interactive option, thereby triggering the above-mentioned target service.

It can be seen that in the related art, when the user selects the above interactive option by "gazing", the length of stay is the sum of T1 and T2.

In this example, in order to predict in advance whether the user is about to trigger the above-mentioned target service when the user selects the above-mentioned interactive option by "gazing", the existing "gazing" mechanism for selecting interactive options can be improved, and the above-mentioned staring waiting process can be omitted, so that the prediction result can be obtained in advance before the user successfully triggers the above-mentioned target service by the traditional "gazing" method.

Specifically, when the user's visual focus enters the area where the above-mentioned interactive option is located, the VR client can still count the length of time the visual focus stays in the area where the interactive option is located in the background; and determine whether the stay time is greater than the preset gaze waiting time threshold T1. If the stay time is greater than T1, it can be directly determined that the above-mentioned target service is about to be triggered by the above-mentioned user.

In this way, the gaze-waiting process in the traditional process of "gazing" to select an interactive option is actually omitted. During the traditional gaze-waiting process, it is possible to predict in advance whether the target service will be triggered by the user.

4) Control of user biometric collection hardware.

In this example, when the VR client successfully predicts that the above-mentioned target service meets the trigger conditions and is about to be triggered by the user according to any of the prediction methods shown above, the VR client can immediately start the above-mentioned biometric collection hardware carried by the VR terminal.

When the biometric feature collection hardware is successfully activated based on the prediction result, the VR client can further confirm the accuracy of the prediction result to determine whether the target service is actually triggered by the user;

For example, the VR client may confirm whether it has received a complete voice command issued by the user for triggering the target service. If so, the VR client may determine that the prediction result is correct and the target service is triggered by the user.

For another example, the VR client may also determine whether the movement trajectory of the user's visual focus passes through the area where the interactive option is located, or determine whether the length of time the user's visual focus stays in the area where the interactive option is located is greater than the sum of T1 and T2 (i.e., gaze waiting and gaze confirmation are completed); if so, the VR client may determine that the prediction result is correct, and the target service is triggered by the user.

That is, in this application, after the VR client completes the prediction of whether the above-mentioned target business meets the triggering conditions, the subsequent triggering process and triggering conditions of the target business can still be consistent with the conventional implementation in the VR scenario.

In this example, if the VR client confirms the prediction result is accurate and the target service is ultimately triggered by the user, the biometric collection hardware will have already been activated. Therefore, the VR client can directly call the biometric collection hardware to collect the user's biometrics and initiate security authentication for the target service based on the collected biometrics. In this case, the user will not experience any delay caused by the hardware initialization of the biometric collection hardware.

When the security authentication for the above-mentioned target business is passed, the VR client can interact with the corresponding server to execute the above-mentioned target business; for example, taking the quick payment business in the VR scene as an example, when the user triggers the quick payment business in the VR scene, the VR client can collect the user's iris features or eye pattern features through the pre-activated biometric camera, and authenticate the user based on the collected iris features or eye pattern features. When the identity authentication is passed, the payment business initiated at this time passes the security authentication, and the VR client can interact with the server to complete the payment.

Of course, if the VR client confirms that the prediction result is inaccurate, and the target service is ultimately not triggered by the user, that is, a prediction error occurs in the prediction process shown above, in this case, the VR client can simply shut down the biometric feature collection hardware that was started in advance.

In this example, the VR client relies on a specific prediction threshold when predicting whether the target service meets the triggering conditions through the various prediction processes shown above;

For example, when predicting based on a user's voice command fragment, the prediction threshold is the similarity threshold described above. When predicting based on the predicted movement trajectory of the user's visual focus from the current moment to N seconds later, the prediction threshold is the value of N (i.e., the preset duration threshold). When predicting based on the length of time the user's visual focus remains in the area where the interactive option is located, the prediction threshold is the gaze wait time threshold described above.

However, due to differences in user behavior, using the same prediction threshold for different users will obviously affect the accuracy of the final prediction result. Therefore, this application also proposes a dynamic prediction threshold mechanism that adjusts the prediction threshold inversely based on the determination result of whether the target service has actually been triggered.

Specifically, when the VR client successfully predicts that the target service meets the triggering condition according to any of the prediction methods shown above, and the VR client determines whether the target service is ultimately triggered by the user:

On the one hand, if it is determined that the target service is ultimately triggered by the user, the prediction threshold used when predicting whether the target service meets the triggering condition can be adjusted to increase the probability of the biometric feature collection hardware being activated;

For example, when the prediction is based on the user's voice command fragment, the above-mentioned prediction threshold is the command similarity threshold described above. In this case, the probability of the above-mentioned biometric feature collection hardware being activated can be increased by lowering the above-mentioned command similarity threshold;

When the prediction is based on the predicted movement trajectory of the user's visual focus from the current moment to N seconds later, the prediction threshold is the value of N (i.e., the preset time threshold). In this case, the probability of the biometric feature collection hardware being activated can be increased by increasing the value of N.

When the prediction is made based on the length of time the user's visual focus stays in the area where the above-mentioned interactive options are located, the above-mentioned prediction threshold is the above-mentioned gaze waiting time threshold. In this case, the probability of the above-mentioned biometric feature collection hardware being activated can be increased by reducing the above-mentioned gaze waiting time threshold.

On the other hand, if the target service is ultimately not triggered by the user, the prediction threshold used when predicting whether the target service meets the triggering condition may be adjusted to reduce the probability of the biometric feature collection hardware being activated.

For example, when the prediction is based on the user's voice command fragment, the above-mentioned prediction threshold is the command similarity threshold described above. In this case, the probability of the above-mentioned biometric feature collection hardware being activated can be reduced by increasing the above-mentioned command similarity threshold;

When the prediction is based on the predicted movement trajectory of the user's visual focus from the current moment to N seconds later, the prediction threshold is the value of N (i.e., the preset time threshold). In this case, the probability of the biometric feature collection hardware being activated can be reduced by reducing the value of N.

When the prediction is made based on the length of time the user's visual focus stays in the area where the above-mentioned interactive options are located, the above-mentioned prediction threshold is the above-mentioned gaze waiting time threshold. In this case, the probability of the above-mentioned biometric feature collection hardware being activated can be reduced by increasing the above-mentioned gaze waiting time threshold.

It should be noted that when adjusting the above-mentioned prediction threshold, the specific increase or decrease is not particularly limited in this application. In actual application, it can be customized based on actual needs.

As can be seen from the description of the above embodiments, the present application uses the collected user operation behavior data to predict whether the target business that requires security authentication based on the user's biometrics meets the trigger condition, and immediately starts the biometric collection hardware when it is predicted that the target business meets the trigger condition;

On the one hand, due to the use of a prediction mechanism for the time when the target service is triggered, the biometric collection hardware can be started in advance when performing security authentication for the above target service, thus ensuring that users do not feel the hardware initialization delay of the biometric collection hardware, thereby improving the user experience;

On the other hand, since the biometric collection hardware remains turned off by default, it will only be turned on when it is predicted that the target business meets the trigger conditions. Therefore, compared with the existing technology, the power consumption of the device can be taken into account as much as possible while ensuring that the user does not feel the hardware initialization delay of the biometric collection hardware.

Corresponding to the above method embodiments, the present application also provides device embodiments.

Please refer to FIG2 , the present application proposes a control device 20 for biometric feature collection hardware, which is applied to a client;

Referring to FIG3 , the hardware architecture of the terminal device serving as the client typically includes a CPU, memory, non-volatile memory, a network interface, and an internal bus. Taking software implementation as an example, the biometric feature acquisition hardware control device 20 can generally be understood as a computer program loaded into memory, which, when executed by the CPU, forms a logical device combining software and hardware. The device 20 includes:

Collection module 201, collects user operation behavior data;

The prediction module 202 predicts whether a target service meets a preset trigger condition based on the collected operation behavior data; wherein the target service is a user service that requires security authentication based on the user's biometric characteristics;

The starting module 203 starts the preset biometric feature collection hardware if it is pre-determined that the target service meets the preset triggering condition.

In this example, the prediction module 202 further:

Determining whether the target service is triggered;

If the target service is triggered, adjusting the prejudgment threshold used when prejudging whether the target service meets a preset triggering condition to increase the probability of the biometric feature collection hardware being activated; and

If the target service is not triggered, the biometric feature collection hardware is turned off, and the pre-judgment threshold used when pre-judging whether the target service meets the preset triggering condition is adjusted to reduce the probability of the biometric feature collection hardware being activated.

In this example, the operation behavior data includes a voice command segment issued by the user; the voice command is used to trigger the target service; the prediction threshold is a preset similarity threshold;

The prediction module 202:

Parse the first voice command segment sent by the user into a corresponding string segment;

Calculating a similarity probability between the character string segment and the instruction character string parsed from the voice instruction, and determining whether the similarity probability is greater than a preset similarity threshold;

If the similarity probability is greater than the similarity threshold, it is pre-determined that the target service meets a preset trigger condition.

In this example, the prediction module 202 further:

If the similarity probability is less than or equal to the similarity threshold, merging the character string segment corresponding to the first voice instruction segment issued by the user with the character string segment corresponding to the next voice instruction segment issued by the user;

Calculating the similarity probability between the merged character string segment and the instruction character string, and re-performing the step of determining whether the similarity probability is greater than the preset similarity threshold;

If the similarity probability is still less than or equal to the similarity threshold, continue to merge the merged instruction string with the next voice instruction segment issued by the user, and re-execute the step of determining whether the similarity probability is greater than the preset similarity threshold until the merged string segment matches the complete instruction string.

In this example, the operation behavior data includes the movement trajectory data of the user's visual focus; the prediction threshold is a preset time threshold;

The prediction module 202:

Inputting the recorded movement trajectory data of the user's visual focus into a preset prediction model for calculation to predict the movement trajectory of the user's visual focus from the current moment to a preset time threshold;

Determining whether the predicted movement trajectory falls within an area where a preset interaction option is located; wherein the preset interaction option is used to trigger the target service;

If the predicted movement trajectory falls into the area where the preset interactive options are located, it is pre-determined that the target service meets the preset triggering condition.

In this example, the operation behavior data includes displacement data of the user's visual focus; the prediction threshold is a preset gaze waiting time threshold;

The prediction module 202:

Determining, based on the recorded movement trajectory data of the user's visual focus, whether the user's visual focus falls within a preset interactive option area;

If it is determined that the user's visual focus falls into the preset interactive option area, the duration of the visual focus staying in the preset interactive option area is counted, and whether the duration is greater than a preset gaze waiting time threshold is determined;

If the dwell time is greater than the gaze waiting time threshold, it is predicted that the target service is about to be triggered by the user.

In this example, the client is a virtual reality client; the biometric feature acquisition hardware is a biometric recognition camera; the biometric feature is a combination of one or more of eye print features, iris features, and palm print features; and the target service is a payment service.

Those skilled in the art will readily appreciate other embodiments of the present application after considering the specification and practicing the invention disclosed herein. This application is intended to cover any variations, uses, or adaptations of the present application that follow the general principles of the present application and include common knowledge or customary techniques in the art not disclosed herein. The description and examples are to be considered as exemplary only, and the true scope and spirit of the present application are indicated by the following claims.

It should be understood that the present application is not limited to the exact structure described above and shown in the drawings, and that various modifications and changes may be made without departing from the scope thereof. The scope of the present application is limited only by the appended claims.

The above description is only a preferred embodiment of the present application and is not intended to limit the present application. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present application shall be included in the scope of protection of the present application.

Claims (12)

1. A control method for biometric data acquisition hardware, characterized in that it is applied to a client, the method comprising: Collect user action data; Based on the collected operational behavior data, it is predicted whether the target service meets the preset triggering conditions; wherein, the target service is a user service that requires security authentication based on the user's biometrics, and the conditions used for prediction are different from the preset triggering conditions; if the target service meets the preset triggering conditions, it indicates that the user is about to trigger the target service; If it is predicted that the target service meets the preset triggering conditions, then the preset biometric collection hardware is activated before the target service is triggered by the user. Determine whether the target service has been triggered; If the target service is triggered, the prediction threshold used to determine whether the target service meets the preset triggering conditions is adjusted to increase the probability that the biometric data acquisition hardware is activated; and, If the target service is not triggered, the biometric data acquisition hardware is turned off, and the prediction threshold used when the target service is predicted to meet the preset triggering conditions is adjusted to reduce the probability that the biometric data acquisition hardware is activated. 2. The method according to claim 1, wherein the operation behavior data includes voice command fragments issued by the user; the voice command is used to trigger the target service; and the prediction threshold is a preset similarity threshold; The method of predicting whether a target service meets preset triggering conditions based on collected operational behavior data includes: The first voice command fragment issued by the user is parsed into a corresponding string fragment; Calculate the similarity probability between the string segment and the instruction string parsed from the voice instruction, and determine whether the similarity probability is greater than a preset similarity threshold; If the similarity probability is greater than the similarity threshold, it is predicted that the target service meets the preset triggering conditions. 3. The method according to claim 2, characterized in that it further comprises: If the similarity probability is less than or equal to the similarity threshold, the string segment corresponding to the first voice command segment issued by the user is merged with the string segment corresponding to the next voice command segment issued by the user. Calculate the similarity probability of the merged string fragments corresponding to the instruction string, and re-determine whether the calculated similarity probability is greater than the preset similarity threshold; If the calculated similarity probability is still less than or equal to the similarity threshold, the merged instruction string is merged with the next voice instruction fragment issued by the user, and the above steps are repeated until the merged string fragment matches the complete instruction string. 4. The method according to claim 1, wherein the operation behavior data includes the movement trajectory data of the user's visual focus; and the prediction threshold is a preset duration threshold; The method of predicting whether a target service meets preset triggering conditions based on collected operational behavior data includes: The recorded data of the user's visual focus movement trajectory is input into a preset prediction model for calculation, so as to predict the movement trajectory of the user's visual focus from the current moment to a preset time threshold. Determine whether the predicted movement trajectory falls within the area of a preset interaction option; wherein the preset interaction option is used to trigger the target service; If the predicted movement trajectory falls within the area of the preset interaction option, it is determined that the target service meets the preset triggering conditions. 5. The method according to claim 1, wherein the operation behavior data includes the displacement data of the user's visual focus; and the prediction threshold is a preset gaze waiting time threshold; The method of predicting whether a target service meets preset triggering conditions based on collected operational behavior data includes: Based on the recorded movement trajectory data of the user's visual focus, it is determined whether the user's visual focus falls within the area of a preset interaction option; wherein, the preset interaction option is used to trigger the target service; If it is determined that the user's visual focus falls on the area where the preset interaction option is located, the duration of the visual focus in the area where the preset interaction option is located is counted, and it is determined whether the duration of the duration of the duration of the duration of the duration of the duration of the duration of the gaze is greater than a preset gaze waiting time threshold. If the dwell time exceeds the gaze waiting time threshold, it is predicted that the target service will be triggered by the user. 6. The method according to claim 1, wherein the client is a virtual reality client; the biometric acquisition hardware is a biometric camera; the biometric feature is one or a combination of iris features, palm print features, and palm print features; and the target business is a payment business. 7. A control device for biometric data acquisition hardware, characterized in that it is applied to a client, the device comprising: The data collection module collects user action data. The prediction module predicts whether a target service meets preset triggering conditions based on the collected operation behavior data; wherein, the target service is a user service that requires security authentication based on the user's biometrics, and the conditions used for prediction are different from the preset triggering conditions; if the target service meets the preset triggering conditions, it indicates that the user is about to trigger the target service; The startup module, if it is predicted that the target service meets the preset triggering conditions, will start the preset biometric collection hardware before the target service is triggered by the user. The prediction module further: Determine whether the target service has been triggered; If the target service is triggered, the prediction threshold used to determine whether the target service meets the preset triggering conditions is adjusted to increase the probability that the biometric data acquisition hardware is activated; and, If the target service is not triggered, the biometric data acquisition hardware is turned off, and the prediction threshold used to determine whether the target service meets the preset triggering conditions is adjusted to reduce the probability that the biometric data acquisition hardware is activated. 8. The apparatus according to claim 7, wherein the operation behavior data includes a voice command segment issued by the user; the voice command is used to trigger the target service; and the prediction threshold is a preset similarity threshold; The prediction module: The first voice command fragment issued by the user is parsed into a corresponding string fragment; Calculate the similarity probability between the string segment and the instruction string parsed from the voice instruction, and determine whether the similarity probability is greater than a preset similarity threshold; If the similarity probability is greater than the similarity threshold, it is predicted that the target service meets the preset triggering conditions. 9. The apparatus according to claim 8, wherein the prediction module further comprises: If the similarity probability is less than or equal to the similarity threshold, the string segment corresponding to the first voice command segment issued by the user is merged with the string segment corresponding to the next voice command segment issued by the user. Calculate the similarity probability of the merged string fragments corresponding to the instruction string, and re-determine whether the calculated similarity probability is greater than the preset similarity threshold; If the calculated similarity probability is still less than or equal to the similarity threshold, the merged instruction string is merged with the next voice instruction fragment issued by the user, and the above steps are repeated until the merged string fragment matches the complete instruction string. 10. The apparatus according to claim 7, wherein the operation behavior data includes the movement trajectory data of the user's visual focus; and the prediction threshold is a preset duration threshold; The prediction module: The recorded data of the user's visual focus movement trajectory is input into a preset prediction model for calculation, so as to predict the movement trajectory of the user's visual focus from the current moment to a preset time threshold. Determine whether the predicted movement trajectory falls within the area of a preset interaction option; wherein the preset interaction option is used to trigger the target service; If the predicted movement trajectory falls within the area of the preset interaction option, it is determined that the target service meets the preset triggering conditions. 11. The apparatus according to claim 7, wherein the operation behavior data includes displacement data of the user's visual focus; and the prediction threshold is a preset gaze waiting time threshold; The prediction module: Based on the recorded movement trajectory data of the user's visual focus, it is determined whether the user's visual focus falls within the area of a preset interaction option; wherein, the preset interaction option is used to trigger the target service; If it is determined that the user's visual focus falls on the area where the preset interaction option is located, the duration of the visual focus in the area where the preset interaction option is located is counted, and it is determined whether the duration of the duration of the duration of the duration of the duration of the duration of the duration of the gaze is greater than a preset gaze waiting time threshold. If the dwell time exceeds the gaze waiting time threshold, it is predicted that the target service will be triggered by the user. 12. The apparatus according to claim 7, wherein the client is a virtual reality client; the biometric acquisition hardware is a biometric camera; the biometric feature is one or a combination of iris features, palm print features, and palm print features; and the target service is a payment service.