Processing method and electronic device

Abstract

The application provides a processing method and an electronic device, and relates to the technical field of artificial intelligence. The processing method comprises the following steps: in response to obtaining a touch signal generated by a target object, acquiring a signal feature of the touch signal; the signal feature at least comprises an intensity feature representing the signal intensity of the touch signal; determining an attribute parameter of the target object according to the signal feature through a target model; determining a processing parameter based on the attribute parameter, and responding to the touch signal generated by the target object based on the processing parameter; wherein the target model comprises a plurality of processing branches, each processing branch has a different network depth; the network depth of a target processing branch used for processing the signal feature is negatively correlated with the signal intensity of the touch signal.

Description

Technical Field

[0001] This application relates to the field of artificial intelligence technology, and in particular to a processing method and an electronic device. Background Technology

[0002] In touchscreen applications, users may wear gloves or use other target objects to touch the screen. The touch signals generated by different target objects are different. Related technologies cannot adaptively adjust the processing method of touch signals according to the characteristics of touch signals. As a result, when the target object changes, the technology cannot adapt to the changes in touch signals, which affects the user's interactive experience. Summary of the Invention

[0003] In view of this, this application provides a processing method and an electronic device.

[0004] According to a first aspect of this application, a processing method is provided, comprising: responding to a touch signal generated by a target object, acquiring signal features of the touch signal; the signal features include at least intensity features characterizing the signal strength of the touch signal; determining attribute parameters of the target object based on the signal features using a target model; determining processing parameters based on the attribute parameters, so as to respond to the touch signal generated by the target object based on the processing parameters; wherein the target model includes multiple processing branches, each processing branch having a different network depth; the network depth of the target processing branch used for processing the signal features is negatively correlated with the signal strength of the touch signal.

[0005] A second aspect of this application provides an electronic device, comprising: a memory for storing computer instructions and a target model; the computer instructions are loaded by a processor for: responding to a touch signal generated by a target object, acquiring signal features of the touch signal; the signal features include intensity features characterizing the signal strength of the touch signal; loading the target model to determine attribute parameters of the target object based on the signal features; and determining processing parameters based on the attribute parameters to respond to the touch signal generated by the target object based on the processing parameters; wherein the target model includes multiple processing branches, each processing branch having a different network depth; the network depth of the target processing branch used for processing the signal features is negatively correlated with the signal strength of the touch signal.

[0006] It should be understood that the description in this section is not intended to identify key or essential features of the embodiments of this application, nor is it intended to limit the scope of this application. Other features of this application will become readily apparent from the following description. Attached Figure Description

[0007] The above and other objects, features and advantages of this application will become clearer from the following description of embodiments with reference to the accompanying drawings, in which:

[0008] Figure 1 A schematic diagram illustrating an example environment in which the methods according to embodiments of this application can be applied;

[0009] Figure 2 The flowchart illustrating one of the processing methods provided in the embodiments of this application is shown in the illustration;

[0010] Figure 3 A schematic diagram of a model structure provided for an embodiment of this application;

[0011] Figure 4 The second flowchart illustrating a processing method provided in an embodiment of this application is shown in the illustration;

[0012] Figure 5 The third flowchart illustrating a processing method provided in an embodiment of this application is shown in the illustration;

[0013] Figure 6 A block diagram of an electronic device provided in an embodiment of this application is shown schematically. Detailed Implementation

[0014] The embodiments of this application will now be described with reference to the accompanying drawings. However, it should be understood that these descriptions are exemplary only and are not intended to limit the scope of this application. In the following detailed description, numerous specific details are set forth to provide a thorough understanding of the embodiments of this application for ease of explanation. However, it will be apparent that one or more embodiments may be implemented without these specific details. Furthermore, descriptions of well-known structures and technologies are omitted in the following description to avoid unnecessarily obscuring the concepts of this application.

[0015] The terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the scope of this application. The terms “comprising,” “including,” etc., as used herein indicate the presence of the stated features, steps, operations, and / or components, but do not exclude the presence or addition of one or more other features, steps, operations, or components.

[0016] All terms used herein (including technical and scientific terms) have the meanings commonly understood by those skilled in the art, unless otherwise defined. It should be noted that the terms used herein are to be interpreted in a manner consistent with the context of this specification, and not in an idealized or overly rigid way.

[0017] When using expressions such as "at least one of A, B, and C", they should generally be interpreted in accordance with the meaning that is commonly understood by those skilled in the art (for example, "a system having at least one of A, B, and C" should include, but is not limited to, a system having A alone, having B alone, having C alone, having A and B, having A and C, having B and C, and / or having A, B, and C, etc.).

[0018] This application provides a processing method and an electronic device. The processing method includes:

[0019] The system responds to touch signals generated by the target object and acquires the signal characteristics of the touch signals; the signal characteristics include at least the intensity characteristics that characterize the signal strength of the touch signals.

[0020] The target object's attribute parameters are determined based on signal characteristics using the target model;

[0021] Processing parameters are determined based on attribute parameters, so as to respond to touch signals generated by the target object based on the processing parameters;

[0022] The target model includes multiple processing branches, each with a different network depth; the network depth of the target processing branch used to process signal features is negatively correlated with the signal strength of the touch signal.

[0023] By adopting the technical solution of this application, when responding to the touch signal generated by the target object, the signal features of the touch signal can be acquired. These signal features include at least the intensity features characterizing the signal strength of the touch signal, thereby perceiving the differences in touch signals generated by different target objects. Then, the target model determines the attribute parameters of the target object based on the signal features, thereby realizing the recognition of the target object. Since the target model includes multiple processing branches, each with a different network depth, and the network depth of the target processing branch used to process signal features is negatively correlated with the signal strength of the touch signal, when the touch signal strength is weak, the processing branch with a deeper network depth is automatically selected to improve the feature extraction capability, and when the touch signal strength is strong, the processing branch with a shallower network depth is automatically selected to ensure processing efficiency, thereby adaptively matching the appropriate processing branch according to the intensity features of the touch signal. Finally, the processing parameters are determined based on the identified attribute parameters, and the touch signal generated by the target object is responded to based on the processing parameters. This realizes the adaptive adjustment of the processing method for the touch signal according to the feature differences of the touch signal, which can adapt to the changes in the touch signal when the target object changes, thus improving the user's interactive experience.

[0024] Figure 1A schematic diagram of an example environment in which the method according to an embodiment of this application can be applied is shown. In this example environment, application 125 is installed on terminal device 110. User 140 can interact with application 125 via terminal device 110 and / or an attached device of terminal device 110.

[0025] In some embodiments, application 125 can be downloaded and installed on terminal device 110. In some embodiments, application 125 can also be accessed in other ways, such as through a web page. Figure 1 In this environment, in response to the launch of application 125, terminal device 110 can display the interface 150 of application 125.

[0026] In some embodiments, terminal device 110 can communicate with server 130 to provide services to application 125. Terminal device 110 can be any type of mobile terminal, fixed terminal, or portable terminal, including mobile phones, desktop computers, laptop computers, notebook computers, netbook computers, tablet computers, media computers, multimedia tablets, personal communication system devices, personal navigation devices, personal digital assistants, audio / video players, digital cameras / camcorders, television receivers, radio receivers, e-book devices, gaming devices, in-vehicle display devices, industrial control terminals, or any combination thereof, including accessories and peripherals of these devices or any combination thereof. In some embodiments, terminal device 110 can also support any type of user-facing interface. Application 125 can be various types of computing systems / servers capable of providing computing power, including but not limited to mainframes, edge computing nodes, and computing devices in cloud environments.

[0027] In some embodiments, application 125 may provide touch interaction functionality. Application 125 may include an application dedicated to providing touch recognition services, or an application integrated with touch recognition functionality. Although Figure 1 The image shows a single application, but in reality, multiple applications can be installed on the terminal device 110.

[0028] In embodiments of this application, multiple models 160 can be deployed locally on the terminal device 110 or remotely. In the case of remote deployment, the terminal device 110 can directly invoke the model 160, or it can invoke the model 160 via the server 130. Exemplarily, the model 160 may have the ability to recognize target object attributes and adaptively adjust processing parameters. The terminal device 110 can provide an interface 150 that presents interactions with the model 160. In the interface 150, the user 140 can interact with the terminal device 110 by generating touch signals through a target object (e.g., a gloved finger). Optionally, the user 140 can operate the terminal device 110 by wearing different types of target objects in different usage scenarios (e.g., outdoor low-temperature environments, industrial workshops, medical clean rooms).

[0029] In the embodiments of this application, during interaction with user 140, model 160 can respond to touch signals generated by user 140, acquire signal characteristics of the touch signals, and determine attribute parameters of the target object. In some embodiments, during signal processing, model 160 can call one or more processing modules 165 to assist in the response processing of touch signals based on the attribute parameters. These processing modules 165 can be of any type, such as signal acquisition modules, feature extraction modules, signal gain adjustment modules, filtering modules, threshold adjustment modules, scanning frequency control modules, etc.

[0030] In some embodiments, the environment may further include a management node for multiple models 160, which can interact with the multiple models 160. In some examples, the management node may, in response to a touch signal from user 140, determine signal features corresponding to the touch signal. The management node may then, based on the signal features, assign the signal features to the model 160 that matches the signal features, requesting that model 160 to determine attribute parameters. In other examples, the management node may also, based on the signal features, determine a signal processing plan. The signal processing plan may indicate one or more processing steps required to complete the signal processing. The management node may assign the one or more processing steps to one or more models 160, which will then execute their respective processing steps. Regarding the management node, in some examples, the management node may be implemented using a target model among the multiple models 160. In other examples, the management node may be implemented using a machine learning model, such as a neural network model.

[0031] In some embodiments, model 160 may be constructed based on one or more machine learning models. In some embodiments, the machine learning model on which model 160 is based may include at least a neural network model, such as a convolutional neural network model. In some embodiments, the machine learning model on which model 160 is based may include a multimodal model capable of processing multiple modal inputs, such as touch signal input, visual input (e.g., an image of the target object), trajectory input, etc. These machine learning models may include feature extraction models capable of generating corresponding outputs based on model inputs. In some embodiments, the machine learning model may receive signal features of touch signals as model inputs and, based on the model inputs, obtain corresponding attribute parameters as model outputs, thereby completing the identification of the target object and the determination of processing parameters.

[0032] It should be understood that the structure and function of the various elements in the environment are described for illustrative purposes only and are not intended to limit the scope of this application in any way.

[0033] The following will be based on Figure 1 The following describes the processing method of the embodiments of this application in detail, based on the described scenario.

[0034] Figure 2 One of the flowcharts illustrating a processing method provided in an embodiment of this application is shown.

[0035] like Figure 2 As shown, the processing method may specifically include the following operations.

[0036] Operation S210 responds to the acquisition of a touch signal generated by the target object and obtains the signal characteristics of the touch signal; the signal characteristics include at least the intensity characteristics that characterize the signal strength of the touch signal;

[0037] Operation S220 determines the attribute parameters of the target object based on the signal characteristics through the target model; wherein, the target model includes multiple processing branches, each processing branch having a different network depth; the network depth of the target processing branch used to process signal characteristics is negatively correlated with the signal strength of the touch signal;

[0038] Operation S230 determines processing parameters based on attribute parameters, and responds to touch signals generated by the target object based on the processing parameters.

[0039] In operation S210, the target object refers to the object that comes into contact with or approaches the touch screen to generate a touch signal. It can be understood as the source of action that triggers the touch screen response and is used to input operations onto the touch screen.

[0040] For example, the target object may include, but is not limited to, gloves, bare fingers, styluses, conductive fabrics, objects with conductive layers, etc.

[0041] Different target objects generate different touch signals. For example, the material and thickness of gloves can affect the capacitive coupling effect, resulting in differences in the capacitance change of the touch signal.

[0042] Therefore, the signal characteristics of the touch signal can be obtained to characterize the physical properties of the target object. The signal characteristics refer to the parameters extracted from the touch signal that can reflect the touch behavior, which can be understood as the quantitative representation of the touch signal in different dimensions.

[0043] Optionally, the signal features include at least intensity features characterizing the signal strength of the touch signal.

[0044] For example, the intensity characteristics may include, but are not limited to, at least one of capacitance change intensity, signal-to-noise ratio, signal amplitude, and baseline drift.

[0045] In one feasible implementation, the capacitance change data at the touch point position can be collected in real time by the capacitive sensor array built into the touch screen, and the ratio of the capacitance change to the baseline capacitance value can be calculated as the capacitance change intensity; at the same time, the ratio of the effective signal power to the noise power of the touch signal can be calculated as the signal-to-noise ratio.

[0046] In another feasible implementation, the signal amplitude at multiple time points can be obtained by time-domain sampling of the touch signal, and the peak value, mean value or integral value of the signal amplitude can be calculated as an intensity feature characterizing the signal strength.

[0047] It should be noted that the intensity characteristics are related to the physical properties of the target object, such as material, thickness, and conductivity. For example, touch signals generated by thick gloves usually have lower capacitance change intensity and lower signal-to-noise ratio, while touch signals generated by thin gloves or bare hands have higher intensity characteristic values.

[0048] In operation S220, the target model refers to a neural network model that has been trained to establish a mapping relationship between signal features and target object attributes. It can be understood as a classification or regression model based on deep learning technology, which is used to output the attribute parameters of the target object based on the input signal features.

[0049] For example, the target model may include, but is not limited to, convolutional neural networks, recurrent neural networks, residual networks, or attention-based neural networks.

[0050] It should be noted that this application does not limit the specific network structure and training method of the target model. The focus is on applying the target model with a specific structure to the touch signal processing scenario and adapting touch signals of different signal strengths through a multi-branch structure.

[0051] In this embodiment, the target model includes multiple processing branches, each with a different network depth. A processing branch refers to an independent network path set up in parallel within the target model, where each branch can process signal features independently. Network depth refers to the number of network layers stacked within a processing branch; a higher network depth indicates a greater number of convolutional layers, pooling layers, or fully connected layers.

[0052] Optionally, the network depth of the target processing branch used to process signal features is negatively correlated with the signal strength of the touch signal. When the touch signal strength is weak, a processing branch with a larger network depth is selected for processing; when the touch signal strength is strong, a processing branch with a smaller network depth is selected for processing. The reason for this design is that when the signal strength is weak, the distinguishability of the signal features is low, requiring a deeper network to extract more complex feature representations; while when the signal strength is strong, the signal features are already quite obvious, and a shallower network can complete the recognition, while reducing computational load and improving response speed.

[0053] The attribute parameters of a target object refer to parameters that characterize the physical properties of the target object. For example, attribute parameters may include, but are not limited to, the type, material category, thickness grade, and conductivity grade of the target object.

[0054] Based on the structure of the target model described above, in one feasible implementation, the signal strength of the touch signal can be determined according to the intensity features in the signal features, and a target processing branch can be selected from multiple processing branches according to the signal strength. The signal features are input into the target processing branch for feature encoding, and finally the attribute parameters and their corresponding confidence levels are output through a fully connected layer.

[0055] In another feasible implementation, a segmentation threshold for signal strength can be preset to divide the signal strength into three levels: strong, medium, and weak, corresponding to three processing branches with network depths of shallow, medium, and deep, respectively; when the signal strength belongs to a certain level, the corresponding processing branch is automatically called for processing.

[0056] It should be noted that the target model needs to be trained using touch signal samples generated by different target objects during the training phase. The samples should cover target objects of different materials and thicknesses so that the model can learn the correspondence between signal features and attribute parameters.

[0057] In operation S230, the processing parameters refer to the configuration parameters used to control the touch screen's response to touch signals. They can be understood as adjustment variables that affect the touch response behavior and are used to adjust the touch screen's recognition and response to touch signals.

[0058] For example, the processing parameters may include, but are not limited to, sensitivity parameters, filtering parameters, gain parameters, scanning frequency, and touch threshold. The sensitivity parameter is used to adjust the sensitivity of the response to the touch signal; the filtering parameter is used to filter out noise interference in the touch signal; the gain parameter is used to adjust the amplification factor of the touch signal; the scanning frequency is used to control the frequency of touch signal acquisition; and the touch threshold is used to determine whether the touch signal is valid.

[0059] The process of determining processing parameters based on attribute parameters is essentially about establishing a correspondence between attribute parameters and processing parameters. Since different target objects generate touch signals with different signal characteristics, different processing parameters are needed to achieve accurate touch response. For example, for target objects with weak signal strength, sensitivity and gain parameters need to be increased to enhance signal recognition capabilities; for target objects with strong signal strength, lower sensitivity parameters can be used to avoid false triggering.

[0060] In one feasible implementation, a mapping table between attribute parameters and processing parameters can be pre-established. The mapping table can record the processing parameter combinations corresponding to different attribute parameters. When the system receives the attribute parameters of the target object, it can query the corresponding processing parameters from the mapping table and configure the processing parameters into the controller of the touch screen so that the touch screen responds to subsequent touch signals according to the processing parameters.

[0061] In another feasible implementation, the value of the processing parameter can be calculated based on the attribute parameter using a preset calculation formula or function; for example, the improvement of the sensitivity parameter can be calculated based on the thickness level of the target object, and the filter type and parameter settings of the filtering parameter can be determined based on the material category.

[0062] It should be noted that the processing parameters can be determined by adjusting a single parameter or by combining multiple parameters to achieve the optimal response to touch signals from different target objects. By dynamically adjusting the processing parameters based on the attribute parameters of the target object, the touchscreen can be adapted to different types of target objects, improving the accuracy of touch operations and the user experience.

[0063] By employing the technical solution of the above embodiments, the attribute parameters of the target object can be identified based on the signal characteristics of the touch signal, and the processing parameters can be dynamically adjusted based on the attribute parameters, thereby enabling the touch screen to adapt to different types of target objects. Since the touch signal strength generated by different target objects varies, uniformly using fixed processing parameters will result in the inaccurate identification of target objects with weak signal strength, or the easy occurrence of accidental touches on target objects with strong signal strength. This embodiment, by acquiring the signal characteristics of the touch signal and using the target model to determine the attribute parameters of the target object, and then adjusting the processing parameters based on the attribute parameters, can provide differentiated touch response strategies for different target objects. This ensures the operational sensitivity of target objects with weak signals, such as thick gloves, while avoiding accidental touches on target objects with strong signals, such as thin gloves or bare hands.

[0064] Furthermore, by setting multiple processing branches with different network depths in the target model, and making the network depth of the target processing branch negatively correlated with the signal strength of the touch signal, an appropriate network depth can be selected for processing touch signals of different strengths. For touch signals with strong signal strength, a processing branch with a smaller network depth can achieve accurate recognition, reducing computational load and improving response speed; for touch signals with weak signal strength, a processing branch with a larger network depth can extract deeper feature representations, improving recognition accuracy. This branch selection strategy ensures recognition accuracy while taking into account computational efficiency, avoiding the waste of computational resources and response delay caused by using deep networks for all touch signals.

[0065] In summary, the technical solution of this embodiment enables the touch screen to adaptively recognize and respond to different target objects, improving the accuracy, sensitivity and user experience of touch operation, while optimizing computational efficiency and response speed through a multi-branch network structure.

[0066] Based on the above embodiments, as an optional embodiment, when the target object is a glove, the multiple processing branches of the target model include at least a first processing branch and a second processing branch; the first processing branch has a first network depth, the second processing branch has a second network depth, and the first network depth is less than the second network depth.

[0067] Based on this, the above operation S220 may also include the following operations.

[0068] The intensity characteristics of the response touch signal indicate that the target object is of the first glove type, and the attribute parameters of the target object are determined based on the signal characteristics through the first processing branch;

[0069] The intensity characteristics of the response touch signal indicate that the target object is of the second glove type, and the attribute parameters of the target object are determined by the second processing branch based on the signal characteristics;

[0070] The thickness of the first glove, characterized by the first glove type, is less than the thickness of the second glove, characterized by the second glove type.

[0071] In this embodiment, the first glove type refers to a glove category with relatively small glove thickness, which can be understood as a category identifier for thin gloves. For example, the first glove type may include, but is not limited to, thin knitted gloves, thin cotton yarn gloves, and single-layer cloth gloves.

[0072] Because the first type of glove is thinner, the glove material has a weaker barrier effect on capacitive coupling, resulting in a higher signal strength for the generated touch signal. Correspondingly, the touch signal strength is characterized by a higher capacitance change intensity and a higher signal-to-noise ratio.

[0073] In one feasible implementation, a first intensity threshold can be preset. When the capacitance change intensity in the intensity feature of the touch signal is greater than the first intensity threshold, the target object is determined to be of the first glove type. In response to the determination result, the signal feature is input into the first processing branch for processing. The first processing branch performs feature encoding on the signal feature and outputs attribute parameters.

[0074] In another feasible implementation, the capacitance change intensity and signal-to-noise ratio can be determined simultaneously. When the capacitance change intensity is greater than a first intensity threshold and the signal-to-noise ratio is greater than a first signal-to-noise ratio threshold, the target object is determined to be of the first glove type, and the first processing branch is called for processing.

[0075] Optionally, the first network depth of the first processing branch can be set to include 3 convolutional layers, and the kernel size of the convolutional layers can be 3×3. Since the touch signal intensity generated by the first glove type is high and the signal features are more distinguishable, a shallower network depth is sufficient to complete the identification of attribute parameters, while reducing computational load and improving response speed.

[0076] Similarly, the second glove type refers to a glove category that is relatively thick, and can be understood as a category identifier for thick gloves. For example, the second glove type may include, but is not limited to, thick cotton gloves, leather gloves, industrial protective gloves, and ski gloves.

[0077] Because the second type of glove is thicker, the glove material strongly impedes capacitive coupling, resulting in a lower signal strength for the generated touch signal. Correspondingly, the touch signal strength is characterized by a lower capacitance change intensity and a lower signal-to-noise ratio.

[0078] In one feasible implementation, a second intensity threshold can be preset, which is less than the first intensity threshold. When the capacitance change intensity in the intensity feature of the touch signal is less than the second intensity threshold, the target object is determined to be a second glove type. In response to the determination result, the signal features are input into a second processing branch for processing. The second processing branch performs feature encoding on the signal features and outputs attribute parameters.

[0079] In another feasible implementation, multiple threshold ranges can be set to classify glove types into multiple thickness levels; when the capacitance change intensity falls into the threshold range corresponding to the second glove type, the target object is determined to be the second glove type.

[0080] Optionally, the second network depth of the second processing branch can be set to include 7 convolutional layers, and residual connections can be introduced into the network structure. Since the touch signal intensity generated by the second glove type is low, the signal features have low discriminative power and are easily affected by noise, a deeper network is needed to extract more complex feature representations to improve the accuracy of attribute parameter recognition.

[0081] It should be noted that the first glove thickness and the second glove thickness refer to the material thickness of the glove in the touch direction. This thickness directly affects the capacitive coupling effect between the finger and the touch screen. The greater the glove thickness, the lower the intensity of the capacitance change in the touch signal, and the more severe the signal attenuation.

[0082] It should be noted that by assigning different glove thicknesses to processing branches with different network depths, adaptive feature extraction capabilities can be provided for touch signals of varying strengths. For the first glove type, using the first processing branch with a smaller network depth can reduce computational complexity while ensuring recognition accuracy; for the second glove type, using the second processing branch with a larger network depth can fully extract effective information from the signal features and improve the ability to recognize weak signals.

[0083] It should be noted that the above-described configuration of the processing branches as the first and second processing branches is merely exemplary, and this application does not limit the specific number of processing branches. In some embodiments, depending on actual application requirements, the processing branches of the target model can be divided into more branches, while the glove type can be further refined.

[0084] For example, a third processing branch can be set up, which has a third network depth between the first network depth and the second network depth; correspondingly, a third glove type can be defined, which represents a third glove thickness between the first glove thickness and the second glove thickness, for processing touch signals generated by medium-thickness gloves.

[0085] In other embodiments, the processing branches can be further refined according to the different glove materials, for example, corresponding processing branches can be set for knitted materials, leather materials, and rubber materials. The number of processing branches and the network depth configuration can be adjusted and optimized according to the actual collected glove touch sample data to achieve accurate identification of different glove types.

[0086] By adopting the technical solution of this embodiment, the appropriate processing branch can be automatically selected for processing according to the different thicknesses of gloves. While ensuring recognition accuracy, the calculation efficiency and response speed are optimized, enabling the touch screen to adapt to different types of gloves from thin to thick, and improving the applicability of touch operation in different glove scenarios.

[0087] Based on the above embodiments, as an optional embodiment, the above operation S230 may further include the following operations.

[0088] The response attribute parameter represents the first attribute parameter of the target object, and the response to the touch signal generated by the target object is processed based on the first processing parameter;

[0089] The response attribute parameter represents the second attribute parameter of the target object, and the response to the touch signal generated by the target object is processed based on the second processing parameter;

[0090] The response sensitivity based on the first processing parameter to the touch signal is different from the response sensitivity based on the second processing parameter to the touch signal.

[0091] In this embodiment, response sensitivity refers to how sensitive the touchscreen is to touch signals. The higher the response sensitivity, the more sensitive the touchscreen is to touch signals, and the easier it is to trigger touch operations; the lower the response sensitivity, the less sensitive the touchscreen is to touch signals, and the stronger the touch signal is required to trigger an operation.

[0092] For example, response sensitivity can be achieved by adjusting at least one of the processing parameters, namely sensitivity parameter, gain parameter, touch threshold, and scanning frequency.

[0093] The first attribute parameter and the second attribute parameter refer to different attribute states of the target object. For example, when the target object is a glove, the first attribute parameter can represent the attribute state of the glove being thinner and the material being thinner, while the second attribute parameter can represent the attribute state of the glove being thicker and the material being thicker.

[0094] For example, when the target object is a stylus, the first attribute parameter can characterize the attribute state of good pen tip conductivity and large pen tip contact area, and the second attribute parameter can characterize the attribute state of weak pen tip conductivity and small pen tip contact area.

[0095] For example, when the target object is a conductive fabric, the first attribute parameter can characterize the attribute state of high conductive fiber weaving density and large conductive layer thickness, and the second attribute parameter can characterize the attribute state of low conductive fiber weaving density and small conductive layer thickness.

[0096] For example, when the target object is a bare hand finger, the first attribute parameter can represent the attribute state of the finger skin being moist and having a large contact area, while the second attribute parameter can represent the attribute state of the finger skin being dry and having a small contact area.

[0097] Because the touch signals generated by target objects with different attribute parameters attenuate to varying degrees, different response sensitivities are required to achieve accurate touch response. Target objects with the first attribute parameter generate touch signals with higher intensity; using an excessively high response sensitivity can easily lead to false triggers. Target objects with the second attribute parameter generate touch signals with lower intensity; using an excessively low response sensitivity will prevent users from completing touch operations correctly.

[0098] In one feasible implementation, corresponding processing parameter combinations can be pre-configured for different attribute parameters; the response attribute parameter is represented as a first attribute parameter, the first processing parameter is queried from the configuration table and applied to the controller of the touch screen, and the response sensitivity corresponding to the first processing parameter is relatively low; the response attribute parameter is represented as a second attribute parameter, the second processing parameter is queried from the configuration table and applied to the controller of the touch screen, and the response sensitivity corresponding to the second processing parameter is relatively high.

[0099] In another feasible implementation, the adjustment range of response sensitivity can be calculated based on attribute parameters; specifically, the increase ratio of sensitivity parameters and the decrease ratio of touch threshold can be calculated based on the target object type, material category, thickness grade or conductivity grade characterized by attribute parameters, so as to achieve differentiated adjustment of response sensitivity.

[0100] For example, for a thin glove characterized by a first attribute parameter, the first processing parameter can be set to increase the signal gain by 20% to 30% and decrease the touch threshold by 10% to 15%; for a thick glove characterized by a second attribute parameter, the second processing parameter can be set to increase the signal gain by 50% to 80% and decrease the touch threshold by 20% to 30%.

[0101] By adopting the technical solution of the above embodiments, the response sensitivity can be dynamically adjusted according to the different attribute parameters of the target object, thereby realizing a differentiated touch response strategy. Since the attenuation of touch signals generated by different target objects varies, using a uniform response sensitivity setting would lead to frequent false triggers for target objects with high signal strength, or the inability to properly recognize target objects with low signal strength. This embodiment, by configuring a first processing parameter with lower response sensitivity for target objects with a first attribute parameter and a second processing parameter with higher response sensitivity for target objects with a second attribute parameter, can ensure the operational accuracy of target objects with strong signals, such as thin gloves, highly conductive styluses, and densely woven conductive fabrics, while ensuring the operability of target objects with weak signals, such as thick gloves, highly conductive styluses, and densely woven conductive fabrics.

[0102] Furthermore, by establishing a correspondence between attribute parameters and processing parameters, adaptive response sensitivity can be provided for target objects with different signal attenuation characteristics. For target objects with low signal attenuation, a lower response sensitivity avoids false triggering caused by oversensitivity; for target objects with high signal attenuation, a higher response sensitivity enhances the ability to identify weak signals, avoiding operational failures caused by sluggish response. This differentiated response strategy reduces the false trigger rate while improving the success rate of touch operations.

[0103] The input and output of the target model in the above embodiments have been described above. The target model provided in the embodiments of this application will be described below.

[0104] Please refer to Figure 3 , Figure 3 This is a schematic diagram of a model structure provided for an embodiment of this application.

[0105] like Figure 3 As shown, the target model 300 may include a feature adaptation layer 310, a feature processing layer 320, and a decision layer 330. The feature processing layer 320 may include a first processing branch 321, a second processing branch 322, and a third processing branch 323.

[0106] Based on the above embodiments, as an optional embodiment, the signal features also include spatial features characterizing the spatial distribution characteristics of the touch signal, and temporal features characterizing the temporal variation characteristics of the touch signal.

[0107] Based on this, the feature adaptation layer 310 is used to determine the target processing branch among multiple processing branches according to at least one of the signal features, spatial features, and temporal features.

[0108] In this embodiment, spatial features refer to the spatial distribution pattern of touch signals on the surface of the touch screen, which can be understood as the geometric and distribution characteristics of touch points in spatial dimensions. For example, spatial features may include, but are not limited to, at least one of touch area, touch point shape profile, touch point pressure distribution, and touch point aspect ratio.

[0109] In one feasible implementation, the capacitance change value of each sensor node in the touch area can be detected by a capacitive sensor array, the number of sensor nodes whose capacitance change value exceeds a preset threshold can be counted, and the touch area can be calculated based on the product of the number of nodes and the area of ​​a single node.

[0110] In another feasible implementation, the boundary coordinates of the touch point can be obtained by extracting the contour of the touch area, and the geometric parameters such as the perimeter, roundness, and convex hull area of ​​the shape enclosed by the boundary coordinates can be calculated as shape contour features.

[0111] It should be noted that touch signals generated by different target objects exhibit different spatial characteristics. For example, gloved fingers, due to their larger contact area and blurred edges, typically have a larger touch area than bare fingers, and their shape and outline have lower roundness; while the pen tip of a stylus has a smaller contact area, and its shape and outline are closer to a circle.

[0112] Similarly, temporal characteristics refer to the dynamic properties of touch signals as they change over time, which can be understood as the pattern of change of touch signals in the time dimension. For example, temporal characteristics may include, but are not limited to, at least one of the following: signal change rate, signal duration, signal rise time, signal fall time, and signal stability.

[0113] In one feasible implementation, the touch signal can be sampled in the time domain to obtain the signal amplitude at multiple consecutive time points. The ratio of the difference in signal amplitude at adjacent time points to the time interval is calculated as the signal change rate. At the same time, the time length from the start of signal triggering to the disappearance of signal is calculated as the signal duration.

[0114] In another feasible implementation, the time it takes for the signal to rise from its initial value to its peak value can be calculated as the signal rise time; the time it takes for the signal to fall from its peak value to its initial value can be calculated as the signal fall time; and the amplitude fluctuation variance of the signal during the stable phase can be calculated as a quantitative indicator of signal stability.

[0115] It should be noted that temporal characteristics can reflect the material properties and operation methods of the target object. For example, knitted gloves, due to their soft material and gaps in the knitted structure, have a slower signal change rate and lower signal stability during touch; while leather gloves, with their denser material, have a shorter signal rise time and higher signal stability.

[0116] In this embodiment, the feature adaptation layer 310 is used to receive signal features including intensity features, spatial features and temporal features, and determine the target processing branch from the first processing branch 321, the second processing branch 322 and the third processing branch 323 according to the signal features.

[0117] In one feasible implementation, the feature adaptation layer 310 can determine the target processing branch using a rule-based approach based on threshold judgment. Specifically, the signal strength of the touch signal can be determined based on the capacitance change intensity in the intensity features, and the corresponding processing branch can be selected based on the signal strength; when the capacitance change intensity is greater than a first intensity threshold, the first processing branch 321 is selected as the target processing branch; when the capacitance change intensity is less than a second intensity threshold, the second processing branch 322 is selected as the target processing branch; when the capacitance change intensity is between the first and second intensity thresholds, the third processing branch 323 is selected as the target processing branch.

[0118] Optionally, branch selection can also take into account at least two of the intensity features, spatial features, and temporal features; if the response capacitance change intensity is low and the touch area is large, it is determined that the target object may be a thick glove, and the second processing branch with a larger network depth is selected 322; if the response capacitance change intensity is high and the touch area is small, it is determined that the target object may be a thin glove or a bare hand, and the first processing branch with a smaller network depth is selected 321.

[0119] In another feasible implementation, the feature adaptation layer 310 can be a trained learning network that learns the mapping relationship between signal features and target processing branches, and predicts the target processing branch to be selected based on the signal features. Specifically, the feature adaptation layer 310 can adopt a gated network structure. The gated network receives signal features as input, encodes the signal features through a multi-layer neural network, and outputs the selection probability of each processing branch; the processing branch with the highest probability is selected as the target processing branch.

[0120] For example, a gating network may include an input layer, a hidden layer, and an output layer; the input layer receives a feature vector containing intensity features, spatial features, and temporal features; the hidden layer transforms and encodes the feature vector through a fully connected layer or a convolutional layer; the output layer outputs a probability distribution corresponding to the number of processing branches, and the selection probability of each processing branch is obtained after normalization.

[0121] It should be noted that when the feature adaptation layer 310 is used as a learning network, it needs to be trained during the model training phase. The training process can employ supervised learning, using training samples labeled with glove type and corresponding processing branch to train the feature adaptation layer 310. The network parameters of the feature adaptation layer 310 are optimized using the backpropagation algorithm, enabling it to accurately predict the target processing branch based on signal features. The training method is a standard deep learning technique and will not be elaborated upon further here.

[0122] In another feasible implementation, the feature adaptation layer 310 can determine the target processing branch by calculating the similarity between the signal features and the feature templates of each processing branch. Feature templates can be pre-established for different glove types, and the feature templates record the intensity feature range, spatial feature range, and temporal feature range corresponding to the glove type; the real-time acquired signal features are matched with each feature template, and the distance or similarity between the signal features and each feature template in the feature space is calculated; the processing branch corresponding to the template with the smallest distance or the highest similarity is selected as the target processing branch.

[0123] By adopting the technical solution of the above embodiments, multimodal signal features can be used to identify the attributes of target objects, improving the accuracy and robustness of identification. Since single-dimensional signal features are easily affected by environmental factors and operating methods, relying solely on intensity features for identification may lead to misjudgment. This embodiment simultaneously extracts the intensity, spatial, and temporal features of the touch signal, characterizing the touch signal from three dimensions: signal intensity, spatial distribution, and temporal variation. This more comprehensively reflects the physical attributes and operational characteristics of the target object, reducing the identification error rate caused by fluctuations in a single feature.

[0124] Based on the above embodiments, as an optional embodiment, the target model further includes a decision layer;

[0125] If the signal features include at least two of the following: intensity features, spatial features, and temporal features, the feature adaptation layer is also used to assign weight coefficients to different signal features respectively, and to perform weighted processing on each signal feature according to the weight coefficient of each signal feature to obtain weighted features;

[0126] The target processing branch is used to encode the weighted features to obtain the encoded features;

[0127] The decision layer is used to determine the attribute parameters of the target object based on the encoded features.

[0128] In this embodiment, the feature adaptation layer 310 may include a learnable attention weight matrix, which is used to dynamically adjust the weights of each feature channel in the multimodal signal features based on the attribute parameters of the target object.

[0129] Specifically, the dimension of the attention weight matrix can be set to correspond to the number of feature channels of the signal features; when the signal features include intensity features, spatial features and temporal features, the attention weight matrix can include three weight coefficients, corresponding to the intensity feature channel, spatial feature channel and temporal feature channel, respectively.

[0130] For example, the attention weight matrix can automatically calculate the weight coefficients of each feature channel based on the input signal features. In one specific implementation, the signal features can be input into the attention weight matrix, transformed through a fully connected layer, and output as a weight vector corresponding to the number of feature channels. Then, the weight vector is normalized to obtain the weight coefficients corresponding to each signal feature.

[0131] In another feasible implementation, weighting coefficients for different signal features can be preset based on the attribute parameters of the target object. If the target object is a knitted glove, since the signal change rate characteristic of knitted material is relatively significant, a higher weighting coefficient, such as 0.5, can be assigned to the time-domain feature, while weighting coefficients of 0.3 and 0.2 can be assigned to the intensity feature and spatial feature, respectively. If the target object is a leather glove, since the capacitance change intensity characteristic of leather material is relatively stable, a higher weighting coefficient, such as 0.5, can be assigned to the intensity feature, while weighting coefficients of 0.3 and 0.2 can be assigned to the spatial feature and time-domain feature, respectively.

[0132] It should be noted that the weight coefficients in the attention weight matrix can be learned during model training. During training, for touch signal samples from different target objects, the model automatically learns the optimal weight allocation for each signal feature using a backpropagation algorithm. For example, when inputting a touch signal sample from a rubber glove, the model learns that the weight coefficient of the intensity feature should be increased to enhance the feature representation capability of the capacitance change intensity channel; when inputting a touch signal sample from a knitted glove, the model learns that the weight coefficient of the time-domain feature should be increased to enhance the feature representation capability of the signal change rate channel. After training, the attention weight matrix can automatically adjust the weight coefficients according to the input of different signal features, achieving dynamic weight allocation.

[0133] By assigning weight coefficients to different signal features and performing weighted processing, adaptive adjustments to the signal characteristics of different target objects can be made. When a certain type of target object has significant discriminative power in a certain feature dimension, weighted processing can strengthen the expression of that feature dimension and improve recognition accuracy; when a certain feature dimension is affected by noise or has low discriminative power, weighted processing can reduce the influence of that feature dimension and avoid noise features interfering with the recognition results.

[0134] In this embodiment, after the feature adaptation layer 310 determines the target processing branch, the target processing branch encodes the weighted features to obtain the encoded features.

[0135] In one feasible implementation, the target processing branch can employ a convolutional neural network structure to encode weighted features. Specifically, after the weighted features are input into the target processing branch, they are processed sequentially through multiple convolutional layers, activation layers, and pooling layers. Each layer further abstracts and extracts the features; after processing by all network layers in the target processing branch, the encoded features are output.

[0136] In another feasible implementation, the target processing branch can use a residual network structure to encode the weighted features. In the residual network, by introducing cross-layer connections, the weighted features retain their original feature information while being processed by multiple layers of the network, avoiding the gradient vanishing problem caused by deep networks and improving the expressive power of the encoded features.

[0137] In this embodiment, the decision layer 330 is used to receive the encoded features output by the target processing branch and determine the attribute parameters of the target object based on the encoded features.

[0138] In one feasible implementation, the decision layer 330 may include a fully connected layer and a classification layer. The fully connected layer receives encoded features and maps the encoded features to a classification space; the classification layer classifies the mapped features and outputs attribute parameters of the target object. For example, when the attribute parameters include glove type and glove thickness level, the classification layer can output the predicted probability of each glove type and select the glove type with the highest predicted probability as the attribute parameter of the target object.

[0139] In another feasible implementation, the decision layer 330 can simultaneously output attribute parameters and their corresponding confidence scores. The confidence score is used to characterize the reliability of the attribute parameter prediction results. Specifically, the decision layer 330 can input encoded features into a classifier, and the classifier outputs the probability distribution of each attribute category. The category corresponding to the highest probability value is used as the attribute parameter, and the highest probability value itself is used as the confidence score.

[0140] Optionally, after determining the attribute parameters, the decision layer 330 can feed the attribute parameters back to the feature adaptation layer 310 to update the attention weight matrix.

[0141] By employing the technical solution of the above embodiments, differentiated weighting processing can be performed according to the importance of different signal features, improving the accuracy and adaptability of target object recognition. Since different target objects have varying degrees of discriminability across different feature dimensions, applying the same weight to all signal features would lead to important features being overlooked or minor features introducing noise. This embodiment assigns weight coefficients to different signal features through a feature adaptation layer, dynamically adjusting the contribution of each feature dimension according to the characteristics of the target object. This results in features with high recognition contribution receiving greater weight, while features with low recognition contribution or high noise levels have reduced weight, thereby improving the accuracy and robustness of the recognition results.

[0142] The following describes the process of adjusting processing parameters based on input data in practical applications.

[0143] Please refer to Figure 4 , Figure 4 The second flowchart illustrating a processing method provided in an embodiment of this application is shown.

[0144] like Figure 4 As shown, the above processing method may also include the following operations.

[0145] Operate S410 to obtain the confidence level corresponding to the output attribute parameters of the target model;

[0146] If the S420 is operated and the response confidence level is less than the confidence threshold, a prompt message will be displayed on the display interface.

[0147] Operation S430 responds to the input data obtained from the target object based on the prompt information, and generates a first adjustment parameter based on the input data; the input data includes at least one of the image of the target object and the input trajectory;

[0148] Operate S440 to adjust the processing parameters based on the first adjustment parameter.

[0149] In operation S420, if the confidence level of the target model output is less than the confidence level threshold, it indicates that the reliability of the current recognition result is low and there is a risk of misidentification. At this time, the secondary verification process is triggered.

[0150] Prompt information refers to guidance information displayed to the user on the screen to prompt the user to perform interactive operations to assist in identification. For example, prompt information may include text prompts, graphic prompts, or animated prompts. For instance, the prompt information could be "Glove operation detected; please redraw the graphic along the dotted line"; simultaneously, a dotted line with a preset trajectory is displayed on the screen to guide the user in the drawing operation.

[0151] In operation S430, input data refers to the data generated when the user interacts with the system in response to prompts. Input data may include at least one of an image of the target object and an input trajectory. The image of the target object can be acquired by a camera on the device containing the touchscreen; the input trajectory can be obtained by recording the sequence of touch point coordinates of the user's drawing operations on the touchscreen.

[0152] In one feasible implementation, when a user performs a redraw operation on the screen, if their finger is close to the front-facing camera, the system can activate the camera to capture an image of the fingertip of the glove as the image of the target object; at the same time, it can record the user's touch trajectory during the redraw process as the input trajectory.

[0153] The first adjustment parameter refers to a parameter value generated based on the input data to adjust the processing parameters. In one feasible implementation, the input data can be analyzed and processed to extract its features, and the first adjustment parameter can be calculated based on these features.

[0154] In operation S440, the first adjustment parameter can be superimposed on the original processing parameter, or the first adjustment parameter can replace part of the value of the original processing parameter.

[0155] Optionally, after adjusting the processing parameters, the system can store the signal characteristics of the current touch signal, the input data, and the corresponding processing parameters as new samples in the database for subsequent model optimization training, thereby improving the model's accuracy in recognizing similar target objects.

[0156] By adopting the technical solution of the above embodiments, when the confidence level of the target model recognition result is low, additional input data can be obtained through a secondary verification process, thereby improving the accuracy of recognition and adjustment. Since a single touch signal may be affected by environmental interference or user operating habits, resulting in insufficient confidence of the target model recognition result, directly adjusting processing parameters based on low-confidence recognition results may lead to inaccurate adjustments or accidental touches. This embodiment determines the reliability of the recognition result by setting a confidence threshold. When the confidence level is lower than the threshold, secondary verification is triggered, guiding the user to provide additional input data. Based on richer data information, a first adjustment parameter is generated, thereby achieving more precise adjustment of processing parameters, improving the accuracy of touch response and user experience.

[0157] Based on the above embodiments, as an optional embodiment, the above operation S430 may further include the following operations.

[0158] The response input data is represented as the input trajectory of the target object based on a preset trajectory feedback, and the trajectory features of the input trajectory are obtained; the trajectory features include at least one of trajectory continuity features, trajectory stability features, and trajectory velocity features.

[0159] The response input data is represented as an image of the target object, and the image of the target object is identified to obtain image features; the image features include at least one of material features and thickness features;

[0160] A first adjustment parameter is generated based on at least one of trajectory features and image features.

[0161] In this embodiment, the preset trajectory refers to a reference trajectory pre-set by the system to guide the user in drawing operations, which can be understood as a standard path used to verify the characteristics of user operations. For example, the preset trajectory may include at least one of the following: straight line, curve, circle, rectangle, and stroke trajectory of letters or numbers.

[0162] In one feasible implementation, when the prompt information is displayed on the display interface, a dotted line of the preset trajectory or a guiding animation can be displayed simultaneously. The user draws along the preset trajectory, and the system records the input trajectory actually drawn by the user.

[0163] Trajectory continuity features refer to characteristic parameters that characterize the degree of continuity of the input trajectory during the drawing process. In one feasible implementation, trajectory continuity features can be calculated by detecting the absence of touch points in the input trajectory; specifically, the number of touch point acquisition interruptions and the duration of these interruptions can be counted. The more interruptions or the longer the duration of the interruptions, the worse the trajectory continuity.

[0164] It should be noted that the continuity of the input trajectory varies depending on the target object. For example, thick gloves, due to their thicker material, result in weaker capacitive coupling, making it easy for touch signals to be interrupted during the drawing process, leading to poor trajectory continuity; while thin gloves or bare hands generate stronger touch signals, resulting in better trajectory continuity.

[0165] Trajectory stability features refer to characteristic parameters that characterize the degree of deviation of the input trajectory relative to the preset trajectory. In one feasible implementation, the distance deviation between the input trajectory and the preset trajectory can be calculated; specifically, the input trajectory and the preset trajectory can be divided into multiple corresponding sampling points, the distance between the corresponding sampling points can be calculated, and the distances can be statistically analyzed to obtain the average offset distance or the maximum offset distance, which serves as the trajectory stability feature.

[0166] It should be noted that trajectory stability characteristics reflect the target object's ability to control the touchscreen. When the target object is a thick glove, the glove material affects the tactile feedback of the fingers, making it difficult for the user to precisely control the drawing path, resulting in a large deviation of the input trajectory from the preset trajectory. When the target object is a thin glove, the user can control the drawing path better, and the deviation of the input trajectory is smaller.

[0167] Trajectory velocity features refer to characteristic parameters that characterize velocity changes during the drawing of an input trajectory. In one feasible implementation, the instantaneous velocity of each sampling point can be obtained by calculating the ratio of the distance between adjacent touch points in the input trajectory to the time interval; the average velocity, velocity variance, or velocity change rate can be obtained by statistically analyzing the instantaneous velocity as the trajectory velocity feature.

[0168] Similarly, in response to input data represented as an image of a target object, image features can be obtained by image recognition. Image features include at least one of material features and thickness features.

[0169] Material characteristics refer to the feature parameters that characterize the surface material properties of a target object. In one feasible implementation, texture and color analysis of the target object's image can be performed using image recognition algorithms; specifically, texture feature vectors and color distribution features of the image can be extracted, and the material category of the target object, such as knitted material, leather material, rubber material, or cotton material, can be identified through a pre-trained classification model.

[0170] Thickness features refer to feature parameters that characterize the thickness attribute of a target object. In one feasible implementation, the thickness can be estimated by analyzing the edge contours and shadow information in the target object image; specifically, the gray-level gradient of the target object edge in the image can be detected. A larger gradient indicates a clearer edge and a smaller corresponding thickness, while a smaller gradient indicates a blurrier edge and a larger corresponding thickness.

[0171] In this embodiment, a first adjustment parameter is generated based on at least one of trajectory features and image features. The first adjustment parameter is used to adjust at least one of the processing parameters: sensitivity parameter, gain parameter, or touch threshold.

[0172] In one feasible implementation, a mapping relationship can be established between trajectory features and a first adjustment parameter. Specifically, when the trajectory continuity feature indicates that the input trajectory has multiple interruptions, it indicates that the touch signal generated by the target object is weak, and the response sensitivity needs to be improved; accordingly, the first adjustment parameter can be set to increase the adjustment range of the sensitivity parameter and decrease the adjustment range of the touch threshold. When the trajectory stability feature indicates that the input trajectory deviates significantly from the preset trajectory, it indicates that the operation accuracy of the target object is affected by the material, and the response sensitivity also needs to be improved; accordingly, the first adjustment parameter can be set to increase the adjustment range of the gain parameter. When the trajectory speed feature indicates that the average speed of the input trajectory is slow or the speed changes significantly, it indicates that the user needs to control the operation more carefully during the drawing process, reflecting that the target object may be a thick glove, and the response sensitivity needs to be improved.

[0173] In another feasible implementation, a mapping relationship can be established between image features and a first adjustment parameter. Specifically, when the material feature represents a knitted material, since the knitted material has a relatively loose weave and weak capacitive coupling effect, the first adjustment parameter can be set to significantly increase the adjustment range of the sensitivity and gain parameters. When the material feature represents a leather material, since leather is relatively dense but may be thick, the first adjustment parameter can be set to moderately increase the adjustment range of the sensitivity parameter. When the thickness feature represents a large thickness of the target object, since increased thickness leads to enhanced touch signal attenuation, the first adjustment parameter can be set to increase the adjustment range of the sensitivity and gain parameters. When the thickness feature represents a small thickness of the target object, the first adjustment parameter can be set to a small adjustment range or keep the current processing parameters unchanged.

[0174] In another feasible implementation, a first adjustment parameter can be generated by combining trajectory features and image features. Specifically, trajectory features and image features can be used as inputs, and the first adjustment parameter can be calculated using a preset adjustment parameter generation model. The adjustment parameter generation model can employ a regression model or a neural network model to learn the mapping relationship between trajectory features and image features and the adjustment magnitude of the processing parameters. If the trajectory continuity features indicate numerous trajectory interruptions and the image feature representation is thick, the adjustment parameter generation model outputs a larger sensitivity enhancement parameter and a larger gain enhancement parameter as the first adjustment parameter; if the trajectory continuity features indicate a coherent trajectory and the image feature representation is thin, the adjustment parameter generation model outputs a smaller adjustment parameter or no adjustment.

[0175] Optionally, after generating the first adjustment parameter, the first adjustment parameter can be applied to test touch operation to verify the adjustment effect; if the response effect of the test touch operation meets the preset requirements, the first adjustment parameter is formally applied to adjust the processing parameter; if the response effect of the test touch operation does not meet the preset requirements, the first adjustment parameter is further adjusted or the user-provided input data is retried.

[0176] By employing the technical solution described in the above embodiments, accurate first adjustment parameters can be generated based on multi-dimensional input data, improving the accuracy and adaptability of parameter adjustment. Since single-dimensional features may have limitations—relying solely on trajectory features may be influenced by user operating habits, and relying solely on image features may be affected by shooting angle and lighting conditions—this embodiment simultaneously extracts the trajectory continuity features, trajectory stability features, and trajectory velocity features of the input trajectory, as well as the material features and thickness features of the target object image. By characterizing the target object from both dynamic operation and static attribute dimensions, it can more comprehensively reflect the true characteristics of the target object.

[0177] Please refer to Figure 5 , Figure 5 The third flowchart illustrating a processing method provided in an embodiment of this application is shown.

[0178] Based on the above embodiments, as an optional embodiment, the above processing method may further include the following operations.

[0179] Operation S510 responds to the touch signal generated by the target object based on the processing parameters, and determines the success rate and false touch rate of the processed touch signal;

[0180] If the response success rate of S520 is less than the success rate threshold or the false touch rate is greater than the false touch threshold, a second adjustment parameter is generated based on the closed-loop control algorithm.

[0181] Operate S530 to adjust the processing parameters based on the second adjustment parameter.

[0182] In operating the S510, the total number of touch operations performed by the user within a certain time window, and the number of touch operations that successfully triggered a response, can be counted. The ratio of the two is the success rate.

[0183] The false touch rate refers to the proportion of unintended touch operations that are incorrectly identified as valid touches. For example, if a user's palm approaches the screen without intending to touch it, but the system identifies it as a valid touch and triggers a response, it is counted as a false touch.

[0184] In operation S520, if the success rate is less than the success rate threshold, it indicates that the current processing parameters cause some valid touches to be unrecognized, and the touch sensitivity needs to be increased; if the false touch rate is greater than the false touch rate threshold, it indicates that the current processing parameters cause too many invalid touches to be misrecognized, and the touch sensitivity needs to be reduced.

[0185] A closed-loop control algorithm refers to a control algorithm that dynamically adjusts based on the deviation between the actual output and the target value. In this embodiment, the closed-loop control algorithm calculates the adjustment direction and magnitude of the processing parameters based on the difference between the actual values ​​of the success rate and the false trigger rate and the threshold, and generates a second adjustment parameter.

[0186] In one feasible implementation, the closed-loop control algorithm can employ a proportional-integral-derivative (PI-DE) control algorithm. Specifically, the deviations between the success rate and the success rate threshold, as well as the deviations between the false trigger rate and the false trigger rate threshold, are calculated and used as inputs to the control algorithm. The control algorithm calculates an adjustment amount based on the current value, historical cumulative value, and rate of change of the deviations, which serves as the second adjustment parameter. If the response success rate is lower than the success rate threshold and the deviation is large, the second adjustment parameter is set to significantly increase the sensitivity parameter; if the response false trigger rate is higher than the false trigger rate threshold and the deviation is large, the second adjustment parameter is set to significantly decrease the sensitivity parameter.

[0187] In operation S530, the second adjustment parameter can be superimposed or multiplied with the current processing parameter to obtain the adjusted processing parameter.

[0188] Optionally, after the processing parameters are adjusted, the success rate and false touch rate of subsequent touch operations can be monitored. If the adjustment effect still does not meet the requirements, a new second adjustment parameter is generated and the processing parameters are adjusted in a further iterative manner to form a closed-loop feedback adjustment process until the success rate and false touch rate both meet the preset requirements.

[0189] It should be noted that the first and second adjustment parameters can work together to adjust the processing parameters. The first adjustment parameter performs feedforward adjustment based on the input data, enabling rapid response to changes in the target object's attributes; the second adjustment parameter performs closed-loop adjustment based on touch effect feedback, continuously optimizing the processing parameters according to actual usage results. Using these two adjustment parameters together allows for rapid initialization and continuous optimization of the processing parameters, improving the accuracy and stability of touch response.

[0190] By adopting the technical solution of the above embodiments, processing parameters can be continuously optimized based on feedback from actual touch effects, thereby improving the accuracy and stability of touch response. Since the actual attributes of the target object may deviate from the model recognition results, or user operating habits and environmental conditions may change, relying solely on one-time parameter adjustments is insufficient to guarantee long-term stable touch performance. This embodiment monitors the success rate and false touch rate of touch operations, and dynamically generates a second adjustment parameter based on actual effect feedback through a closed-loop control algorithm, achieving adaptive optimization of processing parameters.

[0191] Based on the above embodiments, as an optional embodiment, the above processing method may further include the following operations.

[0192] Determine the target application interface for which the touch signal is applied; determine the third adjustment parameter based on the target application interface; adjust the processing parameters based on the third adjustment.

[0193] In this embodiment, the target application interface refers to the application interface currently receiving the touch signal, which can be understood as the application scenario that the user is operating.

[0194] In one feasible implementation, the target application interface can be determined by obtaining the foreground application identifier of the operating system. Different application interfaces have different requirements for touch operation accuracy and response speed. For example, game interfaces typically require fast touch response and high sensitivity to meet the user's real-time operation needs; while text input interfaces require high touch accuracy and low false touch rate to avoid input errors caused by false touches.

[0195] The third adjustment parameter refers to the parameter value used to adjust the processing parameters, determined based on the characteristics of the target application interface. In one feasible implementation, corresponding third adjustment parameters can be pre-set for different types of application interfaces.

[0196] For example, a mapping table between application interface types and third adjustment parameters can be established; if the target application interface is a game interface, the corresponding third adjustment parameter is obtained by querying the mapping table, which is used to improve response speed and sensitivity; if the target application interface is a text input interface, the corresponding third adjustment parameter is obtained by querying the mapping table, which is used to improve touch accuracy and reduce accidental touch rate.

[0197] In another feasible implementation, a third adjustment parameter can be calculated based on historical touch data of the target application interface. Specifically, the historical touch operation characteristics of the user on the application interface are statistically analyzed, including touch frequency, touch area distribution, average touch pressure, etc.; the appropriate adjustment amount of the processing parameters for the application interface is calculated based on the historical touch characteristics, and this is used as the third adjustment parameter.

[0198] After determining the third adjustment parameter, the processing parameters are adjusted based on the third adjustment parameter to adapt to the operational requirements of the current application interface. Specifically, the third adjustment parameter can be superimposed or replaced with the current processing parameters to obtain the adjusted processing parameters.

[0199] Optionally, when a user switches application interfaces, the target application interface can be redefined and the third adjustment parameters can be updated accordingly, so as to realize the dynamic switching of processing parameters according to the application scenario.

[0200] By adopting the technical solution of the above embodiments, processing parameters can be dynamically adjusted according to the characteristics of different application interfaces, thereby improving the adaptability of touch response and user experience. This embodiment identifies the target application interface affected by the touch signal, determines a third adjustment parameter based on the characteristics of the application interface, and adjusts the processing parameters accordingly. This achieves adaptive switching of processing parameters based on application scenarios, enabling touch response to better match the operational needs of different applications and improving the system's scenario adaptability.

[0201] This application also provides an electronic device, including:

[0202] Memory, used to store computer instructions and target models;

[0203] Computer instructions are loaded by the processor for:

[0204] The system responds to touch signals generated by the target object and acquires the signal characteristics of the touch signals; the signal characteristics include intensity characteristics that characterize the signal strength of the touch signals.

[0205] Load the target model to determine the attribute parameters of the target object based on signal characteristics;

[0206] Processing parameters are determined based on attribute parameters, so as to respond to touch signals generated by the target object based on the processing parameters;

[0207] The target model includes multiple processing branches, each with a different network depth; the network depth of the target processing branch used to process signal features is negatively correlated with the signal strength of the touch signal.

[0208] Figure 6 A block diagram of an electronic device provided in an embodiment of the present disclosure is shown schematically. Figure 6 The electronic device shown is merely an example and should not be construed as limiting the functionality and scope of the embodiments disclosed herein.

[0209] like Figure 6 As shown, an electronic device 600 according to an embodiment of this disclosure includes a display screen (not shown) and a processor 601, which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 602 or a program loaded from a memory 608 into a random access memory (RAM) 603. The processor 601 may include, for example, a general-purpose microprocessor (e.g., a CPU), an instruction set processor and / or an associated chipset and / or a special-purpose microprocessor (e.g., an application-specific integrated circuit (ASIC)), etc. The processor 601 may also include onboard memory for caching purposes. The processor 601 may include a single processing unit or multiple processing units for performing different actions of the method flow according to an embodiment of this disclosure.

[0210] RAM 603 stores various programs and data required for the operation of electronic device 600. Processor 601, ROM 602, and RAM 603 are interconnected via bus 604. Processor 601 performs various operations of the method flow according to embodiments of the present disclosure by executing programs in ROM 602 and / or RAM 603. It should be noted that the programs may also be stored in one or more memories other than ROM 602 and RAM 603. Processor 601 may also perform various operations of the method flow according to embodiments of the present disclosure by executing programs stored in said one or more memories.

[0211] According to embodiments of this disclosure, the electronic device 600 may further include an input / output (I / O) interface 605, which is also connected to a bus 604. The system 600 may also include one or more of the following components connected to the input / output (I / O) interface 605: an input device 606 including a keyboard, mouse, etc.; an output device 607 including a cathode ray tube (CRT), liquid crystal display (LCD), display screen, etc., and a speaker, etc.; a memory 608 including a hard disk, etc.; and a communication section 609 including a network interface card such as a LAN card, modem, etc. The communication section 609 performs communication processing via a network such as the Internet. A drive 610 is also connected to the input / output (I / O) interface 605 as needed. A removable medium 611, such as a disk, optical disk, magneto-optical disk, semiconductor memory, etc., is installed on the drive 610 as needed so that computer programs read from it can be installed into the memory 608 as needed.

[0212] According to embodiments of this disclosure, the method flow according to embodiments of this disclosure can be implemented as a computer software program. For example, embodiments of this disclosure include a computer program product comprising a computer program carried on a computer-readable storage medium, the computer program containing program code for performing the methods shown in the flowchart. In such embodiments, the computer program can be downloaded and installed from a network via communication section 609, and / or installed from removable medium 611. When the computer program is executed by processor 601, it performs the functions defined in the system of embodiments of this disclosure. According to embodiments of this disclosure, the systems, devices, apparatuses, modules, units, etc., described above can be implemented by computer program modules.

[0213] This disclosure also provides a computer-readable storage medium, which may be included in the device / apparatus / system described in the above embodiments; or it may exist independently and not assembled into the device / apparatus / system. The computer-readable storage medium carries one or more programs that, when executed, implement the method according to the embodiments of this disclosure.

[0214] According to embodiments of this disclosure, the computer-readable storage medium can be a non-volatile computer-readable storage medium. Examples include, but are not limited to: portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof. In this disclosure, the computer-readable storage medium can be any tangible medium that contains or stores a program that can be used by or in conjunction with an instruction execution system, apparatus, or device.

[0215] For example, according to embodiments of this disclosure, a computer-readable storage medium may include the ROM 602 and / or RAM 603 described above and / or one or more memories other than ROM 602 and RAM 603.

[0216] Embodiments of this disclosure also include a computer program product comprising a computer program containing program code for performing the methods provided in the embodiments of this disclosure. When the computer program product is run on an electronic device, the program code is used to enable the electronic device to implement the methods provided in the embodiments of this disclosure.

[0217] When the computer program is executed by the processor 601, it performs the functions defined in the system / apparatus of this disclosure embodiments. According to embodiments of this disclosure, the systems, apparatuses, modules, units, etc., described above can be implemented by computer program modules.

[0218] In one embodiment, the computer program may rely on a tangible storage medium such as an optical storage device or a magnetic storage device. In another embodiment, the computer program may also be transmitted and distributed in the form of signals over a network medium, and downloaded and installed via the communication section 609, and / or installed from the removable medium 611. The program code contained in the computer program can be transmitted using any suitable network medium, including but not limited to: wireless, wired, etc., or any suitable combination thereof.

[0219] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of the present disclosure. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in a block diagram or flowchart, and combinations of blocks in a block diagram or flowchart, may be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions. Those skilled in the art will understand that the features recited in the various embodiments and / or claims of this disclosure can be combined and / or combined in various ways, even if such combinations or combinations are not expressly described in this disclosure. In particular, the features described in the various embodiments and / or claims of this disclosure may be combined and / or combined in various ways without departing from the spirit and teachings of this disclosure. All such combinations and / or combinations fall within the scope of this disclosure.

[0220] The embodiments of this disclosure have been described above. However, these embodiments are for illustrative purposes only and are not intended to limit the scope of this disclosure. Although various embodiments have been described above, this does not mean that the measures in the various embodiments cannot be used advantageously in combination. The scope of this disclosure is defined by the appended claims and their equivalents. Various substitutions and modifications can be made by those skilled in the art without departing from the scope of this disclosure, and all such substitutions and modifications should fall within the scope of this disclosure.

[0221] The embodiments of this application have been described above. However, these embodiments are merely illustrative and not intended to limit the scope of this application. Although various embodiments have been described above, this does not mean that the measures in the various embodiments cannot be used advantageously in combination. The scope of this application is defined by the appended claims and their equivalents. Various substitutions and modifications can be made by those skilled in the art without departing from the scope of this application, and all such substitutions and modifications should fall within the scope of this application.

Claims

1. A processing method, comprising: In response to a touch signal generated by a target object, the system acquires the signal characteristics of the touch signal; the signal characteristics include at least an intensity characteristic representing the signal strength of the touch signal. The target model determines the attribute parameters of the target object based on the signal characteristics; Processing parameters are determined based on the attribute parameters, so as to respond to the touch signals generated by the target object based on the processing parameters; The target model includes multiple processing branches, each with a different network depth; the network depth of the target processing branch used to process the signal features is negatively correlated with the signal strength of the touch signal.

2. The method according to claim 1, wherein the target object includes gloves; The multiple processing branches of the target model include at least a first processing branch and a second processing branch; The first processing branch has a first network depth, the second processing branch has a second network depth, and the first network depth is less than the second network depth. The step of determining the attribute parameters of the target object based on the signal features using the target model includes: The intensity characteristics of the response to the touch signal characterize the target object as a first glove type, and the attribute parameters of the target object are determined by the first processing branch based on the signal characteristics; The intensity characteristics of the response to the touch signal characterize the target object as a second glove type, and the second processing branch determines the attribute parameters of the target object based on the signal characteristics; Wherein, the thickness of the first glove type is less than the thickness of the second glove type.

3. The method according to claim 1, wherein determining processing parameters based on the attribute parameters to respond to touch signals generated by the target object based on the processing parameters comprises: The first attribute parameter, which represents the target object, is used to process the touch signal generated by the target object based on the first processing parameter. The second attribute parameter, which represents the target object, is used to process the touch signal generated by the target object based on the second processing parameter. The response sensitivity based on the first processing parameter to the touch signal is different from the response sensitivity based on the second processing parameter to the touch signal.

4. The method according to claim 1, wherein the signal features further include spatial features characterizing the spatial distribution characteristics of the touch signal, and temporal features characterizing the temporal variation characteristics of the touch signal; The target model includes a feature adaptation layer; The feature adaptation layer is used to determine the target processing branch among multiple processing branches based on at least one of the signal features' intensity features, spatial features, and temporal features.

5. The method according to claim 4, wherein the target model further comprises a decision layer; If the signal features include at least two of the intensity features, spatial features, and temporal features, the feature adaptation layer is further used to assign weight coefficients to different signal features respectively, and to perform weighted processing on each signal feature according to the weight coefficient of each signal feature to obtain weighted features; The target processing branch is used to encode the weighted features to obtain encoded features; The decision layer is used to determine the attribute parameters of the target object based on the encoded features.

6. The method according to claim 1, further comprising: Obtain the confidence level corresponding to the output attribute parameters of the target model; If the confidence level is less than the confidence threshold, a prompt message will be displayed on the display interface; The system receives input data from the target object based on feedback information and generates a first adjustment parameter based on the input data; the input data includes at least one of an image of the target object and an input trajectory. The processing parameters are adjusted based on the first adjustment parameter.

7. The method according to claim 6, wherein generating the first adjustment parameter based on the input data comprises: The input data is represented as an input trajectory of the target object based on the preset trajectory feedback, and the trajectory features of the input trajectory are obtained; The trajectory features include at least one of trajectory continuity features, trajectory stability features, and trajectory velocity features. In response to the input data representing an image of a target object, image features are obtained by recognizing the image of the target object; the image features include at least one of material features and thickness features; A first adjustment parameter is generated based on at least one of the trajectory features and the image features.

8. The method according to claim 1, further comprising: The system receives touch signals generated by the target object based on the processing parameters and determines the success rate and false touch rate of the processed touch signals. If the success rate is less than the success rate threshold or the false touch rate is greater than the false touch threshold, a second adjustment parameter is generated based on the closed-loop control algorithm; The processing parameters are adjusted based on the second adjustment parameter.

9. The method according to claim 1, further comprising: Determine the target application interface to which the touch signal is applied; The third adjustment parameter is determined based on the target application interface; The processing parameters are adjusted based on the third adjustment.

10. An electronic device, comprising: Memory, used to store computer instructions and target models; The computer instructions are loaded by the processor and used for: In response to a touch signal generated by a target object, the system acquires the signal characteristics of the touch signal; the signal characteristics include intensity characteristics that characterize the signal strength of the touch signal. Load the target model to determine the attribute parameters of the target object based on the signal characteristics using the target model; Processing parameters are determined based on the attribute parameters, so as to respond to the touch signals generated by the target object based on the processing parameters; The target model includes multiple processing branches, each with a different network depth; the network depth of the target processing branch used to process the signal features is negatively correlated with the signal strength of the touch signal.

Images