CN111376910B - User behavior identification method and system and computer equipment - Google Patents
User behavior identification method and system and computer equipment Download PDFInfo
- Publication number
- CN111376910B CN111376910B CN201811635244.1A CN201811635244A CN111376910B CN 111376910 B CN111376910 B CN 111376910B CN 201811635244 A CN201811635244 A CN 201811635244A CN 111376910 B CN111376910 B CN 111376910B
- Authority
- CN
- China
- Prior art keywords
- driving data
- data
- target
- user behavior
- target terminal
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Classifications
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B60—VEHICLES IN GENERAL
- B60W—CONJOINT CONTROL OF VEHICLE SUB-UNITS OF DIFFERENT TYPE OR DIFFERENT FUNCTION; CONTROL SYSTEMS SPECIALLY ADAPTED FOR HYBRID VEHICLES; ROAD VEHICLE DRIVE CONTROL SYSTEMS FOR PURPOSES NOT RELATED TO THE CONTROL OF A PARTICULAR SUB-UNIT
- B60W40/00—Estimation or calculation of non-directly measurable driving parameters for road vehicle drive control systems not related to the control of a particular sub unit, e.g. by using mathematical models
- B60W40/08—Estimation or calculation of non-directly measurable driving parameters for road vehicle drive control systems not related to the control of a particular sub unit, e.g. by using mathematical models related to drivers or passengers
- B60W40/09—Driving style or behaviour
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
Abstract
The embodiment of the disclosure provides a user behavior identification method, an identification system, a computer device and a computer readable storage medium, wherein the user behavior identification method comprises the following steps: acquiring target driving data of a target terminal detected by a target terminal sensor; and analyzing the target driving data through the user behavior recognition model so as to recognize whether the user corresponding to the target terminal has dangerous driving behaviors. The target driving data of the target terminal is analyzed through the user behavior recognition model, and then the driving dangerous operation behaviors of the user carrying the target terminal, such as sudden deceleration, sudden turning and the like, can be accurately and timely judged without configuring additional equipment for the user, so that the personal safety of the user and passengers is guaranteed, and the service quality is effectively improved.
Description
Technical Field
The embodiment of the disclosure relates to the technical field of data processing, in particular to a user behavior identification method, a user behavior identification system, computer equipment and a computer readable storage medium.
Background
Since dangerous driving behaviors of users are important causes of traffic accidents, how to effectively infer dangerous driving behaviors of users is a problem to be solved in order to improve driving safety.
There is a method of analyzing a specific dangerous driving behavior based on a terminal sensor in the related art, but this method has the following problems: 1. the timeliness of data analysis treatment is not enough. Post analysis of large amounts of data is required for data noise reduction and statistical analysis. 2. Erroneous determination of driving behavior in an abnormal scene. For example, sensors deployed at some terminals may not return valid data with high accuracy and low latency, causing the system to determine that the human shaking device is suddenly slowed down. 3. The amount of data is insufficient. Because terminal sensor data more or less have the difference, can lead to the sensor data that same former rule triggered on different terminals to have the variety, only utilize the leading trigger data training model of gathering on some terminal model, often can make the model fall into the predicament of not fitting, can accurately discern dangerous driving identification data on some terminal model promptly, nevertheless be difficult to reach the effect on some terminal.
Disclosure of Invention
The disclosed embodiments are directed to solving at least one of the technical problems of the related art or the related art.
To this end, an aspect of the embodiments of the present disclosure is to provide a method for identifying a user behavior.
Another aspect of the embodiments of the present disclosure is to provide a system for recognizing user behavior.
It is yet another aspect of an embodiment of the present disclosure to provide a computer apparatus.
It is yet another aspect of an embodiment of the present disclosure to provide a computer-readable storage medium.
In view of this, according to an aspect of the embodiments of the present disclosure, a method for identifying a user behavior is provided, where the method includes: acquiring target driving data of a target terminal detected by a target terminal sensor; and analyzing the target driving data through the user behavior recognition model so as to recognize whether the user corresponding to the target terminal has dangerous driving behaviors.
According to the user behavior identification method provided by the embodiment of the disclosure, the target driving data of the target terminal is identified through the user behavior identification model, so that the driving dangerous operation behaviors, such as sudden deceleration, sudden turning and the like, of the user carrying the target terminal are deduced, the dangerous driving behaviors of the user can be accurately and timely judged without configuring additional equipment for the user, the personal safety of the user and passengers is guaranteed, and the service quality is effectively improved.
The method for identifying the user behavior according to the embodiment of the present disclosure may further have the following technical features:
in the above technical solution, preferably, the method further includes: constructing a characteristic learning model, and generating a plurality of virtual driving data by the characteristic learning model; and constructing a user behavior recognition model according to the virtual driving data.
According to the technical scheme, a large amount of virtual driving data are obtained through the feature learning model, the user behavior recognition model is established according to the virtual driving data, the user behavior recognition model which is more accurate in recognition and wider in usability is established through the large amount of virtual driving data, modeling limitation caused by less data is reduced, and the user behavior recognition model is applied to more terminals.
In any of the above technical solutions, preferably, the step of constructing the feature learning model and generating a plurality of virtual driving data from the feature learning model specifically includes: under the same motion behavior condition, acquiring experimental driving data respectively detected by a plurality of terminal sensors, and comparing to obtain difference characteristics among the experimental driving data; training a feature learning model according to the difference features, and generating a plurality of virtual driving data through the feature learning model; wherein the number of virtual driving data is greater than the number of experimental driving data.
In the technical scheme, because terminal sensor data are different more or less, sensor data triggered on different terminals by the same former rule have diversity, and the model is trained by only utilizing experimental driving data collected on part of terminal models, so that the model is often trapped in the difficulty of under-fitting. Therefore, the experimental driving data of the terminals are obtained, the feature learning model is trained by analyzing the difference features between the experimental data, the virtual driving data which is infinitely close to the real driving data is generated, and the user behavior recognition model can independently learn the difference of the terminal data by learning the virtual driving data, so that the model is more robust, and an ideal effect can be achieved on most terminals.
In any of the above technical solutions, preferably, the feature learning model is a GAN (generic adaptive Networks, generating countermeasure network) model; training a feature learning model according to the difference features, and generating a plurality of virtual driving data through the feature learning model, wherein the steps specifically comprise: generating random variables by a generator of the GAN model; judging whether the random variable is the same as the difference characteristic or not through a discriminator of the GAN model so that a generator of the GAN model generates the random variable close to the difference characteristic; and obtaining virtual driving data according to the random variable close to the difference characteristic.
In the technical scheme, the GAN model can generate virtual driving data infinitely close to real driving data by introducing random errors, the specific process is that a generator generates random variables, a judger judges the difference between the random data and difference characteristics, the generator is informed to enable the random data generated again by the generator to be closer to the difference characteristics, a large amount of random data infinitely close to the difference characteristics are finally obtained, a large amount of virtual driving data infinitely close to the real driving data are further obtained, and the fitting performance of the model is improved.
In any one of the above technical solutions, preferably, after acquiring target driving data of the target terminal detected by the target terminal sensor, the method further includes: and carrying out format filtering and acceleration fluctuation filtering on the target driving data.
According to the technical scheme, format filtering is carried out on the target driving data so as to facilitate the identification of the data by a subsequent user behavior identification model, and format errors are avoided. And performing acceleration fluctuation filtering on the target driving data to avoid misjudgment of driving behaviors in abnormal scenes. For example, sensors deployed at some terminals may not return valid data with high accuracy and low latency, causing the system to determine that the human shaking device is suddenly slowed down.
In any of the above technical solutions, preferably, the method further includes: when the user corresponding to the target terminal is identified to have dangerous driving behaviors, recording target driving data; and sending alarm information to the target terminal.
In the technical scheme, when the user carrying the target terminal is identified to have dangerous driving behaviors, the target driving data and the dangerous driving behaviors are recorded so as to directly judge the dangerousness when the behaviors occur again, and meanwhile, alarm information is sent to inform the terminal user that the dangerous driving behaviors occur.
In any of the above technical solutions, preferably, the experimental driving data, the virtual driving data, and the target driving data at least include: acceleration data, included angle data, speed data and duration data.
In the technical scheme, the experimental driving data, the virtual driving data and the target driving data at least comprise but are not limited to: the method comprises the steps of establishing an accurate user behavior recognition model according to acceleration data, included angle data, speed data and duration data of experimental driving data and virtual driving data, and recognizing the data of target driving data through the user behavior recognition model.
According to another aspect of the embodiments of the present disclosure, a system for identifying user behavior is provided, the system including: the system comprises an acquisition unit, a display unit and a display unit, wherein the acquisition unit is used for acquiring target driving data of a target terminal detected by a target terminal sensor; and the identification unit is used for analyzing the target driving data through the user behavior identification model so as to identify whether the user corresponding to the target terminal has dangerous driving behaviors.
According to the user behavior identification system provided by the embodiment of the disclosure, the target driving data of the target terminal is identified through the user behavior identification model, so that the driving dangerous operation behaviors, such as sudden deceleration, sudden turning and the like, of the user carrying the target terminal are deduced, the dangerous driving behaviors of the user can be accurately and timely judged without configuring additional equipment for the user, the personal safety of the user and passengers is guaranteed, and the service quality is effectively improved.
The system for identifying the user behavior according to the embodiment of the present disclosure may further have the following technical features:
in the above technical solution, preferably, the method further includes: the data generating unit is used for constructing a characteristic learning model and generating a plurality of virtual driving data from the characteristic learning model; and the training unit is used for constructing a user behavior recognition model according to the virtual driving data.
According to the technical scheme, a large amount of virtual driving data are obtained through the feature learning model, the user behavior recognition model is established according to the virtual driving data, the user behavior recognition model which is more accurate in recognition and wider in usability is established through the large amount of virtual driving data, modeling limitation caused by less data is reduced, and the user behavior recognition model is applied to more terminals.
In any one of the above technical solutions, preferably, the data generating unit includes: the comparison unit is used for acquiring experimental driving data respectively detected by the plurality of terminal sensors under the same motion behavior condition and comparing to obtain difference characteristics among the experimental driving data; the data generation unit is specifically used for training the feature learning model according to the difference features and generating a plurality of virtual driving data through the feature learning model; wherein the number of virtual driving data is greater than the number of experimental driving data.
In the technical scheme, because terminal sensor data are different more or less, sensor data triggered on different terminals by the same former rule have diversity, and the model is trained by only utilizing experimental driving data collected on part of terminal models, so that the model is often trapped in the difficulty of under-fitting. Therefore, the experimental driving data of the terminals are obtained, the feature learning model is trained by analyzing the difference features between the experimental data, the virtual driving data which is infinitely close to the real trigger data is generated, and the user behavior recognition model can independently learn the difference of the terminal data by learning the virtual driving data, so that the model is more robust, and an ideal effect can be achieved on most terminals.
In any of the above technical solutions, preferably, the feature learning model generates a countermeasure network GAN model; a data generation unit, specifically configured to generate a random variable by a generator of the GAN model; judging whether the random variable is the same as the difference characteristic or not through a discriminator of the GAN model so that a generator of the GAN model generates the random variable close to the difference characteristic; and obtaining virtual driving data according to the random variable close to the difference characteristic.
In the technical scheme, the GAN model can generate virtual driving data infinitely close to real driving data by introducing random errors, the specific process is that a generator generates random variables, a judger judges the difference between the random data and difference characteristics, the generator is informed to enable the random data generated again by the generator to be closer to the difference characteristics, a large amount of random data infinitely close to the difference characteristics are finally obtained, a large amount of virtual driving data infinitely close to the real driving data are further obtained, and the fitting performance of the model is improved.
In any of the above technical solutions, preferably, the method further includes: and the data filtering unit is used for performing format filtering and acceleration fluctuation filtering on the target driving data after acquiring the target driving data of the target terminal detected by the target terminal sensor.
According to the technical scheme, format filtering is carried out on the target driving data so as to facilitate the identification of the data by a subsequent user behavior identification model, and format errors are avoided. And performing acceleration fluctuation filtering on the target driving data to avoid misjudgment of driving behaviors in abnormal scenes. For example, sensors deployed at some terminals may not return valid data with high accuracy and low latency, causing the system to determine that the human shaking device is suddenly slowed down.
In any of the above technical solutions, preferably, the method further includes: the recording unit is used for recording target driving data when the user corresponding to the target terminal is identified to have dangerous driving behaviors; and the sending unit is used for sending the alarm information to the target terminal.
In the technical scheme, when the user carrying the target terminal is identified to have dangerous driving behaviors, the target driving data and the dangerous driving behaviors are recorded so as to directly judge the dangerousness when the behaviors occur again, and meanwhile, alarm information is sent to inform the terminal user that the dangerous driving behaviors occur.
In any of the above technical solutions, preferably, the experimental driving data, the virtual driving data, and the target driving data at least include: acceleration data, included angle data, speed data and duration data.
In the technical scheme, the experimental driving data, the virtual driving data and the target driving data at least comprise but are not limited to: the method comprises the steps of establishing an accurate user behavior recognition model according to acceleration data, included angle data, speed data and duration data of experimental driving data and virtual driving data, and recognizing the data of target driving data through the user behavior recognition model.
According to a further aspect of the embodiments of the present disclosure, a computer device is provided, which includes a memory, a processor and a computer program stored on the memory and executable on the processor, wherein the processor implements the steps of the method for identifying a user behavior according to any one of the above items when executing the computer program.
The computer device provided by the embodiment of the disclosure realizes all the advantages of the method for identifying the user behavior as described in any one of the above when the processor executes the computer program.
According to yet another aspect of an embodiment of the present disclosure, a computer-readable storage medium is proposed, on which a computer program is stored, which, when being executed by a processor, carries out the steps of the method for identifying a user behavior according to any one of the above.
The computer-readable storage medium provided by the embodiment of the present disclosure, when being executed by a processor, realizes all the advantages of the method for identifying a user behavior as described in any one of the above.
Additional aspects and advantages of the disclosed embodiments will be set forth in part in the description which follows or may be learned by practice of the disclosed embodiments.
Drawings
The above and/or additional aspects and advantages of the embodiments of the present disclosure will become apparent and readily appreciated from the following description of the embodiments, taken in conjunction with the accompanying drawings of which:
FIG. 1 shows a flow diagram of a method of identifying user behavior of one embodiment of the present disclosure;
FIG. 2 shows a flow diagram of a method of identifying user behavior of another embodiment of the present disclosure;
FIG. 3 illustrates a flow diagram of a method of identifying user behavior according to yet another embodiment of the disclosure;
FIG. 4 illustrates a GAN model schematic diagram of one embodiment of the disclosed embodiments;
FIG. 5 shows a flow diagram of a method of identifying user behavior of yet another embodiment of the present disclosure;
FIG. 6 is a diagram illustrating a method for identifying user behavior in accordance with one embodiment of the present disclosure;
FIG. 7 shows a schematic block diagram of a recognition system of user behavior of one embodiment of the present disclosure;
FIG. 8 shows a schematic block diagram of a recognition system of user behavior of another embodiment of the disclosed embodiments;
FIG. 9 shows a schematic block diagram of a recognition system of user behavior of yet another embodiment of the disclosed embodiments;
FIG. 10 shows a schematic block diagram of a recognition system of user behavior of yet another embodiment of the disclosed embodiments;
FIG. 11 shows a schematic block diagram of a computer device of one embodiment of the disclosed embodiments.
Detailed Description
In order that the above objects, features and advantages of the embodiments of the present disclosure can be more clearly understood, embodiments of the present disclosure will be described in further detail below with reference to the accompanying drawings and detailed description. It should be noted that the embodiments and features of the embodiments of the present application may be combined with each other without conflict.
In the following description, numerous specific details are set forth in order to provide a thorough understanding of the embodiments of the disclosure, however, the embodiments of the disclosure may be practiced in other ways than those described herein, and therefore the scope of the embodiments of the disclosure is not limited to the specific embodiments disclosed below.
In an embodiment of the first aspect of the embodiment of the present disclosure, a method for identifying a user behavior is provided, and fig. 1 illustrates a flowchart of the method for identifying a user behavior according to an embodiment of the present disclosure. Wherein, the method comprises the following steps:
102, acquiring target driving data of a target terminal detected by a target terminal sensor;
and 104, analyzing the target driving data through the user behavior recognition model to recognize whether the user corresponding to the target terminal has dangerous driving behaviors.
According to the user behavior identification method provided by the embodiment of the disclosure, the target driving data of the target terminal is identified through the user behavior identification model, so that the driving dangerous operation behaviors, such as sudden deceleration, sudden turning and the like, of the user carrying the target terminal are deduced, the dangerous driving behaviors of the user can be accurately and timely judged without configuring additional equipment for the user, the personal safety of the user and passengers is guaranteed, and the service quality is effectively improved.
Fig. 2 shows a flowchart of a user behavior recognition method according to another embodiment of the present disclosure. Wherein, the method comprises the following steps:
and step 208, analyzing the target driving data through the user behavior recognition model to recognize whether the user corresponding to the target terminal has dangerous driving behaviors.
In the embodiment, a large amount of virtual driving data is obtained through the feature learning model, the user behavior recognition model is established according to the virtual driving data, the user behavior recognition model which is more accurate in recognition and wider in usability is established through the large amount of virtual driving data, modeling limitation caused by less data is reduced, and the user behavior recognition model is applied to more terminals.
Fig. 3 is a flowchart illustrating a user behavior recognition method according to still another embodiment of the present disclosure. Wherein, the method comprises the following steps:
and 310, analyzing the target driving data through the user behavior recognition model to recognize whether the user corresponding to the target terminal has dangerous driving behaviors.
In this embodiment, as the terminal sensor data are different more or less, the sensor data triggered on different terminals by the same former rule have diversity, and the model is trained only by using the experimental driving data collected on part of terminal models, which often causes the model to fall into the difficulty of under-fitting. Therefore, the experimental driving data of the terminals are obtained, the feature learning model is trained by analyzing the difference features between the experimental data, the virtual driving data which is infinitely close to the real driving data is generated, and the user behavior recognition model can independently learn the difference of the terminal data by learning the virtual driving data, so that the model is more robust, and an ideal effect can be achieved on most terminals.
In one embodiment of the disclosed embodiment, preferably, the feature learning model is a GAN model; training a feature learning model according to the difference features, and generating a plurality of virtual driving data through the feature learning model, wherein the steps specifically comprise: generating random variables by a generator of the GAN model; judging whether the random variable is the same as the difference characteristic or not through a discriminator of the GAN model so that a generator of the GAN model generates the random variable close to the difference characteristic; and obtaining virtual driving data according to the random variable close to the difference characteristic.
In this embodiment, the GAN model may generate virtual driving data infinitely close to real driving data by introducing a random error, and as shown in fig. 4, a random variable is generated by the generator, a judgment unit judges a difference between the random data and the difference feature, the generator is informed to make the random data generated again by the generator closer to the difference feature, so as to obtain a large amount of random data infinitely close to the difference feature, further obtain a large amount of virtual driving data infinitely close to the real driving data, and further improve the fitting performance of the model.
Fig. 5 shows a flowchart of a method for identifying user behavior according to another embodiment of the disclosure. Wherein, the method comprises the following steps:
In the embodiment, after the target driving data of the target terminal detected by the target terminal sensor is obtained, format filtering is performed on the target driving data so as to facilitate the identification of the data by a subsequent user behavior identification model, and a format error is avoided. And performing acceleration fluctuation filtering on the target driving data to avoid misjudgment of driving behaviors in abnormal scenes. For example, sensors deployed at some terminals may not return valid data with high accuracy and low latency, causing the system to determine that the human shaking device is suddenly slowed down.
When the dangerous driving behavior of the user carrying the target terminal is recognized, the target driving data and the dangerous driving behavior are recorded so as to directly judge the danger when the behavior appears again, and meanwhile, alarm information is sent to inform the terminal user that the dangerous driving behavior occurs.
In one embodiment of the present disclosure, preferably, the experimental driving data, the virtual driving data, and the target driving data each include at least: acceleration data, included angle data, speed data and duration data.
In this embodiment, the experimental driving data, the virtual driving data, and the target driving data each include at least, but are not limited to: the method comprises the steps of establishing an accurate user behavior recognition model according to acceleration data, included angle data, speed data and duration data of experimental driving data and virtual driving data, and recognizing the data of target driving data through the user behavior recognition model.
Fig. 6 is a schematic diagram illustrating a method for identifying user behavior according to a specific embodiment of the present disclosure. Wherein, the method comprises the following steps: the method comprises the steps of collecting driving data of the mobile device, sequentially carrying out format filtering, acceleration fluctuation filtering and shaking two classification model filtering on the driving data, carrying out behavior recognition on the driving data to a user behavior recognition model based on a GAN model through a transfer server and a decision server, recording the driving data if the driving data is recognized as dangerous driving behaviors, and sending alarm information to the mobile device.
In a second aspect of the embodiments of the present disclosure, a system for identifying user behavior is provided, and fig. 7 shows a schematic block diagram of a system 700 for identifying user behavior according to an embodiment of the present disclosure. Among other things, the system 700 includes:
an acquisition unit 702 configured to acquire target driving data of a target terminal detected by a target terminal sensor; and the identifying unit 704 is used for analyzing the target driving data through the user behavior identification model so as to identify whether the user corresponding to the target terminal has dangerous driving behaviors.
The system 700 for identifying user behaviors provided by the embodiment of the present disclosure identifies target driving data of a target terminal through a user behavior identification model, and then deduces driving dangerous operation behaviors, such as sudden deceleration, sudden turning, and the like, of a user carrying the target terminal, and can accurately and timely judge the dangerous driving behaviors of the user without configuring additional equipment for the user, thereby ensuring personal safety of the user and passengers, and effectively improving service quality.
Fig. 8 shows a schematic block diagram of a recognition system 800 of user behavior of another embodiment of the disclosed embodiments. Among other things, the system 800 includes:
a data generation unit 802 configured to construct a feature learning model, and generate a plurality of virtual driving data from the feature learning model; the training unit 804 is used for constructing a user behavior recognition model according to the virtual driving data; an obtaining unit 806, configured to obtain target driving data of a target terminal detected by a target terminal sensor; the identifying unit 808 is configured to analyze the target driving data through the user behavior identification model to identify whether the user corresponding to the target terminal has dangerous driving behavior.
In the embodiment, a large amount of virtual driving data is obtained through the feature learning model, the user behavior recognition model is established according to the virtual driving data, the user behavior recognition model which is more accurate in recognition and wider in usability is established through the large amount of virtual driving data, modeling limitation caused by less data is reduced, and the user behavior recognition model is applied to more terminals.
Fig. 9 shows a schematic block diagram of a recognition system 900 of user behavior of yet another embodiment of the disclosed embodiments. Among other things, the system 900 includes:
the data generation unit 902 includes: the comparison unit 922 is configured to obtain experimental driving data respectively detected by the multiple terminal sensors under the same motion behavior condition, and compare the experimental driving data to obtain a difference characteristic between the experimental driving data; the feature learning model is a countermeasure network GAN model;
a data generation unit 902, specifically configured to generate a random variable by a generator of the GAN model; judging whether the random variable is the same as the difference characteristic or not through a discriminator of the GAN model so that a generator of the GAN model generates the random variable close to the difference characteristic; obtaining virtual driving data according to the random variable close to the difference characteristic; the number of the virtual driving data is larger than that of the experimental driving data;
a training unit 904, configured to construct a user behavior recognition model according to the virtual driving data;
an acquisition unit 906 configured to acquire target driving data of a target terminal detected by a target terminal sensor;
the identifying unit 908 is configured to analyze the target driving data through the user behavior identification model to identify whether a user corresponding to the target terminal has dangerous driving behavior.
In this embodiment, as the terminal sensor data are different more or less, the sensor data triggered on different terminals by the same former rule have diversity, and the model is trained only by using the experimental driving data collected on part of terminal models, which often causes the model to fall into the difficulty of under-fitting. Therefore, the experimental driving data of the terminals are obtained, the feature learning model is trained by analyzing the difference features between the experimental data, the virtual driving data which is infinitely close to the real trigger data is generated, and the user behavior recognition model can independently learn the difference of the terminal data by learning the virtual driving data, so that the model is more robust, and an ideal effect can be achieved on most terminals.
The GAN model can generate virtual driving data infinitely close to real driving data by introducing random errors, the specific process is that random variables are generated by a generator, a judging device judges the difference size of the random data and the difference characteristics, the generator is informed to enable the random data generated by the generator again to be closer to the difference characteristics, a large amount of random data infinitely close to the difference characteristics are finally obtained, a large amount of virtual driving data infinitely close to the real driving data are further obtained, and the fitting performance of the model is improved.
Fig. 10 shows a schematic block diagram of a recognition system 100 of user behavior of yet another embodiment of the disclosed embodiments. Wherein the system 100 comprises:
the data generation unit 102 includes: a comparing unit 1022, configured to obtain experimental driving data respectively detected by multiple terminal sensors under the same motion behavior condition, and compare the experimental driving data to obtain a difference characteristic between the experimental driving data; the feature learning model is a countermeasure network GAN model;
a data generation unit 102, specifically configured to generate a random variable by a generator of the GAN model; judging whether the random variable is the same as the difference characteristic or not through a discriminator of the GAN model so that a generator of the GAN model generates the random variable close to the difference characteristic; obtaining virtual driving data according to the random variable close to the difference characteristic; the number of the virtual driving data is larger than that of the experimental driving data;
the training unit 104 is used for constructing a user behavior recognition model according to the virtual driving data;
an acquisition unit 106, configured to acquire target driving data of a target terminal detected by a target terminal sensor;
the data filtering unit 108 is configured to perform format filtering and acceleration fluctuation filtering on the target driving data after acquiring the target driving data of the target terminal detected by the target terminal sensor;
the identification unit 110 is configured to analyze target driving data through a user behavior identification model to identify whether a user corresponding to a target terminal has dangerous driving behavior;
the recording unit 112 is used for recording target driving data when the user corresponding to the target terminal is identified to have dangerous driving behaviors;
a sending unit 114, configured to send the alarm information to the target terminal.
In the embodiment, the target driving data is subjected to format filtering so as to be convenient for the subsequent user behavior recognition model to recognize the data, and the occurrence of format errors is avoided. And performing acceleration fluctuation filtering on the target driving data to avoid misjudgment of driving behaviors in abnormal scenes. For example, sensors deployed at some terminals may not return valid data with high accuracy and low latency, causing the system to determine that the human shaking device is suddenly slowed down.
When the dangerous driving behavior of the user carrying the target terminal is recognized, the target driving data and the dangerous driving behavior are recorded so as to directly judge the danger when the behavior appears again, and meanwhile, alarm information is sent to inform the terminal user that the dangerous driving behavior occurs.
In one embodiment of the present disclosure, preferably, the experimental driving data, the virtual driving data, and the target driving data each include at least: acceleration data, included angle data, speed data and duration data.
In this embodiment, the experimental driving data, the virtual driving data, and the target driving data each include at least, but are not limited to: the method comprises the steps of establishing an accurate user behavior recognition model according to acceleration data, included angle data, speed data and duration data of experimental driving data and virtual driving data, and recognizing the data of target driving data through the user behavior recognition model.
In an embodiment of the third aspect of the embodiments of the present disclosure, a computer device is provided, and fig. 11 shows a schematic block diagram of a computer device 120 according to an embodiment of the present disclosure. Wherein the computer device 120 comprises:
a memory 122, a processor 124 and a computer program stored on the memory 122 and executable on the processor 124, the processor 124 implementing the steps of the method for identifying user behavior as any one of the above when executing the computer program.
The computer device 120 provided by the embodiment of the present disclosure, when the processor 124 executes the computer program, achieves all the advantages of the method for identifying a user behavior as described in any one of the above.
An embodiment of the fourth aspect of the embodiments of the present disclosure provides a computer-readable storage medium, on which a computer program is stored, where the computer program, when executed by a processor, implements the steps of the method for identifying a user behavior according to any one of the above.
The computer-readable storage medium provided by the embodiment of the present disclosure, when being executed by a processor, realizes all the advantages of the method for identifying a user behavior as described in any one of the above.
In the description herein, reference to the term "one embodiment," "some embodiments," "a specific embodiment," or the like, means that a particular feature, structure, material, or characteristic described in connection with the embodiment or example is included in at least one embodiment or example of the embodiments of the disclosure. In this specification, the schematic representations of the terms used above do not necessarily refer to the same embodiment or example. Furthermore, the particular features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples.
The above description is only a preferred embodiment of the present disclosure and is not intended to limit the present disclosure, and various modifications and changes may be made to the present disclosure by those skilled in the art. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the disclosed embodiments should be included in the scope of protection of the disclosed embodiments.
Claims (14)
1. A method for identifying user behavior, the method comprising:
constructing a feature learning model based on difference features, generating a plurality of virtual driving data by the feature learning model, wherein the difference features are experimental driving data respectively detected by sensors arranged on a plurality of terminals under the same motion behavior condition, and comparing to obtain the difference features among the experimental driving data;
constructing the user behavior recognition model according to the virtual driving data;
acquiring target driving data of a target terminal detected by a target terminal sensor;
and analyzing the target driving data through a user behavior recognition model so as to recognize whether a user corresponding to the target terminal has dangerous driving behaviors.
2. The method for identifying user behavior according to claim 1,
the number of the virtual driving data is greater than the number of the experimental driving data.
3. The method according to claim 2, wherein the feature learning model is a countermeasure network GAN model;
training the feature learning model according to the difference features, and generating a plurality of virtual driving data through the feature learning model, specifically including:
generating random variables by a generator of the GAN model;
judging whether the random variable is the same as the difference feature through a discriminator of the GAN model so that a generator of the GAN model generates the random variable close to the difference feature;
and obtaining the virtual driving data according to the random variable close to the difference characteristic.
4. The method for recognizing user behavior according to any one of claims 1 to 3, further comprising, after acquiring the target driving data of the target terminal detected by the target terminal sensor:
and carrying out format filtering and acceleration fluctuation filtering on the target driving data.
5. The method for identifying user behavior according to any one of claims 1 to 3, further comprising:
when the dangerous driving behavior of the user corresponding to the target terminal is identified, recording the target driving data;
and sending alarm information to the target terminal.
6. The method according to claim 2, wherein the experimental driving data, the virtual driving data, and the target driving data each include at least: acceleration data, included angle data, speed data and duration data.
7. A system for identifying user behavior, the system comprising:
the data generating unit is used for constructing a feature learning model based on the difference features between experimental driving data respectively detected by a plurality of terminal sensors under the same motion behavior condition, and generating a plurality of virtual driving data by the feature learning model;
the training unit is used for constructing the user behavior recognition model according to the virtual driving data;
the system comprises an acquisition unit, a display unit and a display unit, wherein the acquisition unit is used for acquiring target driving data of a target terminal detected by a target terminal sensor;
the identification unit is used for analyzing the target driving data through a user behavior identification model so as to identify whether a user corresponding to the target terminal has dangerous driving behaviors;
the data generation unit includes:
and the comparison unit is used for acquiring experimental driving data respectively detected by the plurality of terminal sensors under the same motion behavior condition and comparing to obtain the difference characteristics among the experimental driving data.
8. The system for recognizing user behavior according to claim 7,
the number of the virtual driving data is greater than the number of the experimental driving data.
9. The system according to claim 8, wherein the feature learning model is a countermeasure network GAN model;
the data generation unit is specifically used for generating random variables by the generator of the GAN model; judging whether the random variable is the same as the difference feature through a discriminator of the GAN model so that a generator of the GAN model generates the random variable close to the difference feature; and obtaining the virtual driving data according to the random variable close to the difference characteristic.
10. The system according to any one of claims 7 to 9, further comprising:
and the data filtering unit is used for performing format filtering and acceleration fluctuation filtering on the target driving data after acquiring the target driving data of the target terminal detected by the target terminal sensor.
11. The system according to any one of claims 7 to 9, further comprising:
the recording unit is used for recording the target driving data when the user corresponding to the target terminal is identified to have the dangerous driving behavior;
and the sending unit is used for sending alarm information to the target terminal.
12. The system for identifying user behavior according to claim 8, wherein the experimental driving data, the virtual driving data, and the target driving data each include at least: acceleration data, included angle data, speed data and duration data.
13. Computer device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, characterized in that the processor implements the steps of the method for recognition of user behavior according to any one of claims 1 to 6 when executing the computer program.
14. A computer-readable storage medium, on which a computer program is stored, which, when being executed by a processor, carries out the steps of the method for identifying a user behavior according to any one of claims 1 to 6.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811635244.1A CN111376910B (en) | 2018-12-29 | 2018-12-29 | User behavior identification method and system and computer equipment |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811635244.1A CN111376910B (en) | 2018-12-29 | 2018-12-29 | User behavior identification method and system and computer equipment |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111376910A CN111376910A (en) | 2020-07-07 |
| CN111376910B true CN111376910B (en) | 2022-04-15 |
Family
ID=71212998
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811635244.1A Active CN111376910B (en) | 2018-12-29 | 2018-12-29 | User behavior identification method and system and computer equipment |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111376910B (en) |
Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP1498717A2 (en) * | 2003-07-15 | 2005-01-19 | Universität Stuttgart | Procedure and device for determining the crosswind behaviour of a vehicle |
| CN104463244A (en) * | 2014-12-04 | 2015-03-25 | 上海交通大学 | Aberrant driving behavior monitoring and recognizing method and system based on smart mobile terminal |
| CN106934876A (en) * | 2017-03-16 | 2017-07-07 | 广东翼卡车联网服务有限公司 | A kind of recognition methods of vehicle abnormality driving event and system |
| CN108108766A (en) * | 2017-12-28 | 2018-06-01 | 东南大学 | Driving behavior recognition methods and system based on Fusion |
| CN108230426A (en) * | 2018-02-07 | 2018-06-29 | 深圳市唯特视科技有限公司 | A kind of image generating method based on eye gaze data and image data set |
| CN108267172A (en) * | 2018-01-25 | 2018-07-10 | 神华宁夏煤业集团有限责任公司 | Mining intelligent robot inspection system |
| CN109063724A (en) * | 2018-06-12 | 2018-12-21 | 中国科学院深圳先进技术研究院 | A kind of enhanced production confrontation network and target sample recognition methods |
| CN109086668A (en) * | 2018-07-02 | 2018-12-25 | 电子科技大学 | Based on the multiple dimensioned unmanned aerial vehicle remote sensing images road information extracting method for generating confrontation network |
| CN109102014A (en) * | 2018-08-01 | 2018-12-28 | 中国海洋大学 | The image classification method of class imbalance based on depth convolutional neural networks |
Family Cites Families (14)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US6577960B1 (en) * | 2000-07-13 | 2003-06-10 | Simmonds Precision Products, Inc. | Liquid gauging apparatus using a time delay neural network |
| IE20030437A1 (en) * | 2003-06-11 | 2004-12-15 | Scient Systems Res Ltd | A method for process control of semiconductor manufacturing equipment |
| JP2011170856A (en) * | 2010-02-22 | 2011-09-01 | Ailive Inc | System and method for motion recognition using a plurality of sensing streams |
| CN102982336B (en) * | 2011-09-02 | 2015-11-25 | 株式会社理光 | Model of cognition generates method and system |
| JP2013205171A (en) * | 2012-03-28 | 2013-10-07 | Sony Corp | Information processing device, information processing method, and program |
| CN105551182A (en) * | 2015-11-26 | 2016-05-04 | 吉林大学 | Driving state monitoring system based on Kinect human body posture recognition |
| US10474880B2 (en) * | 2017-03-15 | 2019-11-12 | Nec Corporation | Face recognition using larger pose face frontalization |
| US10489992B2 (en) * | 2017-05-08 | 2019-11-26 | Lear Corporation | Vehicle communication network |
| CN107194987B (en) * | 2017-05-12 | 2021-12-10 | 西安蒜泥电子科技有限责任公司 | Method for predicting human body measurement data |
| CN107607182B (en) * | 2017-08-04 | 2019-12-13 | 广西大学 | truck weighing system and weighing method |
| CN108345869B (en) * | 2018-03-09 | 2022-04-08 | 南京理工大学 | Driver posture recognition method based on depth image and virtual data |
| CN108710831B (en) * | 2018-04-24 | 2021-09-21 | 华南理工大学 | Small data set face recognition algorithm based on machine vision |
| CN109002686B (en) * | 2018-04-26 | 2022-04-08 | 浙江工业大学 | Multi-grade chemical process soft measurement modeling method capable of automatically generating samples |
| CN108763857A (en) * | 2018-05-29 | 2018-11-06 | 浙江工业大学 | A kind of process soft-measuring modeling method generating confrontation network based on similarity |
-
2018
- 2018-12-29 CN CN201811635244.1A patent/CN111376910B/en active Active
Patent Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP1498717A2 (en) * | 2003-07-15 | 2005-01-19 | Universität Stuttgart | Procedure and device for determining the crosswind behaviour of a vehicle |
| CN104463244A (en) * | 2014-12-04 | 2015-03-25 | 上海交通大学 | Aberrant driving behavior monitoring and recognizing method and system based on smart mobile terminal |
| CN106934876A (en) * | 2017-03-16 | 2017-07-07 | 广东翼卡车联网服务有限公司 | A kind of recognition methods of vehicle abnormality driving event and system |
| CN108108766A (en) * | 2017-12-28 | 2018-06-01 | 东南大学 | Driving behavior recognition methods and system based on Fusion |
| CN108267172A (en) * | 2018-01-25 | 2018-07-10 | 神华宁夏煤业集团有限责任公司 | Mining intelligent robot inspection system |
| CN108230426A (en) * | 2018-02-07 | 2018-06-29 | 深圳市唯特视科技有限公司 | A kind of image generating method based on eye gaze data and image data set |
| CN109063724A (en) * | 2018-06-12 | 2018-12-21 | 中国科学院深圳先进技术研究院 | A kind of enhanced production confrontation network and target sample recognition methods |
| CN109086668A (en) * | 2018-07-02 | 2018-12-25 | 电子科技大学 | Based on the multiple dimensioned unmanned aerial vehicle remote sensing images road information extracting method for generating confrontation network |
| CN109102014A (en) * | 2018-08-01 | 2018-12-28 | 中国海洋大学 | The image classification method of class imbalance based on depth convolutional neural networks |
Also Published As
| Publication number | Publication date |
|---|---|
| CN111376910A (en) | 2020-07-07 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107154950B (en) | Method and system for detecting log stream abnormity | |
| CN111857356B (en) | Method, device, equipment and storage medium for recognizing interaction gesture | |
| CN108090458B (en) | Human body falling detection method and device | |
| CN109600336B (en) | Verification code application method, device and computer readable storage medium | |
| JP6823501B2 (en) | Anomaly detection device, anomaly detection method and program | |
| CN109413023B (en) | Training of machine recognition model, machine recognition method and device, and electronic equipment | |
| CN104751110A (en) | Bio-assay detection method and device | |
| CN108241580B (en) | Client program testing method and terminal | |
| CN111414813A (en) | Dangerous driving behavior identification method, device, equipment and storage medium | |
| CN110458126B (en) | Pantograph state monitoring method and device | |
| CN107682317B (en) | method for establishing data detection model, data detection method and equipment | |
| CN107729729B (en) | Automatic passing test method of sliding verification code based on random forest | |
| CN111385297A (en) | Wireless device fingerprint identification method, system, device and readable storage medium | |
| CN109426700B (en) | Data processing method, data processing device, storage medium and electronic device | |
| CN112953971A (en) | Network security traffic intrusion detection method and system | |
| CN110808995B (en) | Safety protection method and device | |
| CN109995751B (en) | Internet access equipment marking method and device, storage medium and computer equipment | |
| CN111376910B (en) | User behavior identification method and system and computer equipment | |
| CN112487265A (en) | Data processing method and device, computer storage medium and electronic equipment | |
| CN111478922A (en) | Method, device and equipment for detecting communication of hidden channel | |
| CN116302809A (en) | Edge end data analysis and calculation device | |
| CN116112209A (en) | Vulnerability attack flow detection method and device | |
| CN113361455B (en) | Training method of face counterfeit identification model, related device and computer program product | |
| CN115189961A (en) | Fault identification method, device, equipment and storage medium | |
| CN112929458B (en) | Method and device for determining address of server of APP (application) and storage medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |