In-vehicle living body detection method and device, vehicle and storage medium
By leveraging the vehicle's wireless communication system and AI-trained RF signal feedback models, the method addresses the high cost and low accuracy issues of existing in-vehicle detection systems, achieving improved detection accuracy and reduced hardware needs.
Patent Information
- Application Number
- CN202310807682.6
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2023-07-03
- Publication Date
- 2025-01-03
AI Technical Summary
Existing in-vehicle living body detection systems require high costs and have low accuracy due to the use of separate hardware systems like millimeter-wave radars, infrared, and camera recognition, which are prone to interference and obstruction, leading to blind spots and privacy concerns.
Utilize the vehicle's existing wireless communication system to transmit radio frequency signals for liveness detection, combining them with AI models trained on RF signal feedback data and image data to improve detection accuracy and reduce hardware requirements.
This approach reduces costs by reusing existing systems, enhances detection accuracy through RF signal penetration and diffraction capabilities, and improves detection by avoiding obstacles, thus providing high technical and cost benefits.
Smart Images
Figure CN119247481A_ABST
Abstract
Description
Technical Field
[0001] The present disclosure relates to the field of intelligent driving technology, and in particular to a method and device for detecting a living body in a vehicle, a vehicle, and a storage medium. Background Art
[0002] With the development of science and technology, vehicles are used more and more frequently in people's daily lives. Therefore, detecting vital signs inside vehicles has become one of the important measures to protect people's lives and property. In related technologies, a complete hardware system is set up in the vehicle to detect living bodies inside the vehicle. However, this solution requires high costs and has low accuracy in detecting living bodies inside the vehicle. Summary of the invention
[0003] The present disclosure provides a method and device for detecting a live body in a vehicle, a vehicle and a storage medium, so as to at least solve the technical problems of high cost and low detection accuracy of detecting a live body in a vehicle in the related art. The technical solution of the present disclosure is as follows:
[0004] According to a first aspect of an embodiment of the present disclosure, a method for detecting a living body in a vehicle is provided, which is applied to a vehicle and includes:
[0005] Control the wireless communication system to transmit a wireless radio frequency signal to the interior of the vehicle to obtain radio frequency signal feedback data;
[0006] The target liveness detection model is controlled to perform liveness detection according to the RF signal feedback data to obtain a liveness detection result, wherein the target liveness detection model is trained according to a set of RF signal feedback sample data, and the set of RF signal feedback sample data includes RF signal feedback sample data corresponding to at least one scene inside the vehicle.
[0007] Optionally, before the control target living body detection model performs living body detection according to the radio frequency signal feedback data, the method further includes:
[0008] Acquire an image sample data set and a radio frequency signal feedback sample data set, wherein the image sample data set includes image sample data corresponding to at least one scene inside the vehicle, and the image sample data and the radio frequency signal feedback sample data correspond one to one;
[0009] The initial liveness detection model is trained according to the image sample data set and the radio frequency signal feedback sample data set to obtain a target liveness detection model that meets the training conditions.
[0010] Optionally, the training of the initial liveness detection model according to the image sample data set and the radio frequency signal feedback sample data set to obtain a target liveness detection model that meets the training conditions includes:
[0011] Training an initial liveness detection model according to the image sample data set and the radio frequency signal feedback sample data set to obtain a trained liveness detection model;
[0012] Using a radio frequency signal feedback test data set to test the trained initial liveness detection model to obtain a detection accuracy corresponding to the trained liveness detection model, wherein the radio frequency signal feedback test data set includes radio frequency signal feedback test data corresponding to at least one scene inside the vehicle;
[0013] If the detection accuracy is lower than the accuracy threshold, the trained liveness detection model is used as the initial liveness detection model for retraining until the detection accuracy is not lower than the accuracy threshold, thereby obtaining a target liveness detection model that meets the training conditions.
[0014] Optionally, the training of the initial liveness detection model according to the image sample data set and the radio frequency signal feedback sample data set to obtain the trained liveness detection model includes:
[0015] marking and classifying at least one image sample data in the image sample data set and at least one radio frequency signal feedback sample data in the radio frequency signal feedback sample data set to obtain a marked and classified image sample data set and a marked and classified radio frequency signal feedback sample data set;
[0016] The initial living body detection model is trained according to the labeled and classified image sample data set and the labeled and classified radio frequency signal feedback sample data set to obtain a trained living body detection model.
[0017] Optionally, when the initial living body detection model is trained according to the labeled and classified image sample data set and the labeled and classified radio frequency signal feedback sample data set, at least one of the following training methods is adopted:
[0018] Transfer learning training;
[0019] supervised learning training;
[0020] Distillation learning training.
[0021] Optionally, the training of the initial liveness detection model according to the labeled and classified image sample data set and the labeled and classified radio frequency signal feedback sample data set to obtain the trained liveness detection model includes:
[0022] Training the initial liveness detection model according to the labeled and classified image sample data set to obtain a liveness detection model after image training;
[0023] The liveness detection model trained with the image is trained according to the radio frequency signal feedback sample data set after the label classification to obtain the trained liveness detection model.
[0024] According to a second aspect of an embodiment of the present disclosure, a method and device for detecting a living body in a vehicle is provided, which is applied to a vehicle and includes:
[0025] A signal transceiver unit, used to control the wireless communication system to transmit a wireless radio frequency signal to the interior of the vehicle and obtain radio frequency signal feedback data;
[0026] A liveness detection unit is used to control a target liveness detection model to perform liveness detection according to the RF signal feedback data to obtain a liveness detection result, wherein the target liveness detection model is trained based on a set of RF signal feedback sample data, and the set of RF signal feedback sample data includes RF signal feedback sample data corresponding to at least one scene inside the vehicle.
[0027] Optionally, the device further includes a model training unit, which is used, before the control target liveness detection model performs liveness detection according to the radio frequency signal feedback data, to:
[0028] Acquire an image sample data set and a radio frequency signal feedback sample data set, wherein the image sample data set includes image sample data corresponding to at least one scene inside the vehicle, and the image sample data and the radio frequency signal feedback sample data correspond one to one;
[0029] The initial liveness detection model is trained according to the image sample data set and the radio frequency signal feedback sample data set to obtain a target liveness detection model that meets the training conditions.
[0030] Optionally, the model training unit is used to train an initial liveness detection model according to the image sample data set and the radio frequency signal feedback sample data set to obtain a target liveness detection model that meets the training conditions, specifically for:
[0031] Training an initial liveness detection model according to the image sample data set and the radio frequency signal feedback sample data set to obtain a trained liveness detection model;
[0032] Using a radio frequency signal feedback test data set to test the trained initial liveness detection model to obtain a detection accuracy corresponding to the trained liveness detection model, wherein the radio frequency signal feedback test data set includes radio frequency signal feedback test data corresponding to at least one scene inside the vehicle;
[0033] If the detection accuracy is lower than the accuracy threshold, the trained liveness detection model is used as the initial liveness detection model for retraining until the detection accuracy is not lower than the accuracy threshold, thereby obtaining a target liveness detection model that meets the training conditions.
[0034] Optionally, the model training unit is used to train an initial liveness detection model according to the image sample data set and the radio frequency signal feedback sample data set, and when obtaining the trained liveness detection model, is specifically used to:
[0035] marking and classifying at least one image sample data in the image sample data set and at least one radio frequency signal feedback sample data in the radio frequency signal feedback sample data set to obtain a marked and classified image sample data set and a marked and classified radio frequency signal feedback sample data set;
[0036] The initial living body detection model is trained according to the labeled and classified image sample data set and the labeled and classified radio frequency signal feedback sample data set to obtain a trained living body detection model.
[0037] Optionally, the model training unit is used to train the initial living body detection model according to the labeled and classified image sample data set and the labeled and classified radio frequency signal feedback sample data set, using at least one of the following training methods:
[0038] Transfer learning training;
[0039] supervised learning training;
[0040] Distillation learning training.
[0041] Optionally, the model training unit is used to train the initial liveness detection model according to the labeled and classified image sample data set and the labeled and classified radio frequency signal feedback sample data set, and when the trained liveness detection model is obtained, it is specifically used to:
[0042] Training the initial liveness detection model according to the labeled and classified image sample data set to obtain a liveness detection model after image training;
[0043] The liveness detection model trained with the image is trained according to the radio frequency signal feedback sample data set after the label classification to obtain the trained liveness detection model.
[0044] According to a third aspect of an embodiment of the present disclosure, a vehicle is provided, comprising:
[0045] processor;
[0046] a memory for storing instructions executable by the processor;
[0047] The processor is configured to execute the instructions to implement the in-vehicle living body detection method described in any one of the aforementioned aspects.
[0048] According to a fourth aspect of the present application, a storage medium is provided. When instructions in the storage medium are executed by a processor of a vehicle, the vehicle is enabled to execute the in-vehicle living body detection method described in any one of the aforementioned aspects.
[0049] According to a fifth aspect of the present application, a computer program product is provided, comprising a computer program, wherein the computer program implements any one of the methods described in the preceding aspects when executed by a processor.
[0050] The technical solution provided by the embodiments of the present disclosure brings at least the following beneficial effects:
[0051] 1) By reusing the existing wireless communication system in the vehicle for liveness detection, there is no need to set up a complete hardware system separately, which can reduce the cost of in-vehicle liveness detection;
[0052] 2) The RF signal is rich in information and has strong penetration and diffraction capabilities. It can effectively avoid obstacles and perform blind spot recognition in the entire cockpit, thus improving the accuracy of detection.
[0053] 3) Detection of living things through artificial intelligence network training combined with radio frequency signals can achieve extremely high technical, performance and cost benefits.
[0054] It is to be understood that the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the present disclosure. BRIEF DESCRIPTION OF THE DRAWINGS
[0055] The drawings herein are incorporated into and constitute a part of the specification, illustrate embodiments consistent with the present disclosure, and together with the description are used to explain the principles of the present disclosure, and do not constitute improper limitations on the present disclosure.
[0056] Figure 1 A schematic diagram showing a flow chart of a method for detecting a living body in a vehicle provided by an embodiment of the present disclosure;
[0057] Figure 2 A schematic diagram showing a flow chart of another in-vehicle living body detection method provided by an embodiment of the present disclosure;
[0058] Figure 3 A schematic diagram showing a flow chart of another in-vehicle living body detection method provided by an embodiment of the present disclosure;
[0059] Figure 4 A schematic diagram of the structure of a vehicle liveness detection device provided by an embodiment of the present disclosure is shown;
[0060] Figure 5 is a block diagram of a vehicle according to an exemplary embodiment. DETAILED DESCRIPTION
[0061] In order to enable ordinary persons in the art to better understand the technical solutions of the present disclosure, the technical solutions in the embodiments of the present disclosure will be clearly and completely described below in conjunction with the accompanying drawings.
[0062] It should be noted that the terms "first", "second", etc. in the specification and claims of the present disclosure and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments of the present disclosure described herein can be implemented in an order other than those illustrated or described herein. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the present disclosure. Instead, they are merely examples of devices and methods consistent with some aspects of the present disclosure as detailed in the appended claims.
[0063] In the related art, a complete set of hardware systems is separately set up in the vehicle, such as millimeter wave radar, infrared, ultra wide band technology (Ultra Wide Band, UWB) or camera recognition corresponding hardware systems to detect living things in the vehicle.
[0064] However, millimeter-wave radars are easily blocked by obstacles or seats in the car, resulting in blind spots. Infrared systems can also be blocked by obstacles, and different internal and external heat sources in different situations in the car can interfere with and challenge infrared recognition. Camera recognition systems pose privacy risks, and living things in the car can be easily blocked by seats or objects. UWB systems are easily interfered by the multipath environment in the car, resulting in reduced recognition capabilities.
[0065] In summary, the above solutions have low accuracy in in-vehicle liveness detection and require the addition of additional separate hardware systems, which are costly.
[0066] The present disclosure is described in detail below with reference to specific embodiments.
[0067] In the first embodiment, if Figure 1 As shown, Figure 1 A flow chart of a method for detecting a living body in a vehicle provided by an embodiment of the present disclosure is shown, and the method can be implemented by a computer program and can be run on a device for performing the method for detecting a living body in a vehicle. The method is executed by a vehicle.
[0068] Specifically, the in-vehicle liveness detection method includes:
[0069] S101, controlling the wireless communication system to transmit a wireless radio frequency signal to the interior of the vehicle to obtain radio frequency signal feedback data;
[0070] According to some embodiments, the wireless communication system refers to a system installed inside the vehicle for implementing communication through a wireless protocol. The wireless communication system may be, for example, a wireless network (WIFI) communication system.
[0071] In some embodiments, a wireless radio frequency signal refers to a signal that is transmitted into the air through an antenna by modulating (amplitude modulation or frequency modulation) an electrical information source (analog or digital) with a high-frequency current. The wireless radio frequency signal can be, for example, a frequency modulated continuous wave transmitted by a transmitting module in a WIFI communication system.
[0072] In some embodiments, the RF signal feedback data refers to feedback data generated based on feedback information such as the reflected signal and delay corresponding to the wireless RF signal. The reflected signal can be, for example, a reflected wave received by a receiving module in a WIFI communication system, and the WIFI communication system can de-modulate the reflected wave and restore it to an electrical information source. The feedback data includes, but is not limited to, RF signal field strength distribution data, RF signal field strength distribution diagram, etc.
[0073] It is easy to understand that when a vehicle performs in-vehicle liveness detection, the vehicle can control the wireless communication system to transmit wireless radio frequency signals to the interior of the vehicle to obtain radio frequency signal feedback data.
[0074] S102, controlling the target liveness detection model to perform liveness detection according to the radio frequency signal feedback data to obtain a liveness detection result.
[0075] According to some embodiments, the target liveness detection model refers to a neural network module for in-vehicle liveness detection obtained by training based on a set of radio frequency signal feedback sample data. The target liveness detection model includes but is not limited to a convolutional neural network (CNN) model, a K nearest neighbor (KNN) model, a natural language processing (NLP), etc.
[0076] In some embodiments, the RF signal feedback sample data set includes RF signal feedback sample data corresponding to at least one scene inside the vehicle. The RF signal feedback sample data refers to feedback data obtained as samples after transmitting a wireless RF signal to the inside of the vehicle.
[0077] According to some embodiments, the liveness detection result includes but is not limited to liveness quantity information, liveness location information, etc.
[0078] It is easy to understand that when the vehicle obtains the RF signal feedback data, the vehicle can control the target liveness detection model to perform liveness detection according to the RF signal feedback data to obtain the liveness detection result.
[0079] In summary, the method provided by the embodiment of the present disclosure controls the wireless communication system to transmit wireless radio frequency signals to the interior of the vehicle to obtain radio frequency signal feedback data; controls the target liveness detection model to perform liveness detection according to the radio frequency signal feedback data to obtain liveness detection results. Therefore, by reusing the original wireless communication system in the vehicle for liveness detection, there is no need to set up a complete hardware system separately, and the drawbacks of the existing detection system can be avoided, such as privacy leakage, high dependence on software and hardware performance, and additional costs of independent system configuration. At the same time, the radio frequency signal is rich in information, and has strong penetration and diffraction capabilities. It can effectively avoid the obstruction of obstacles to perform blind spot recognition in the entire cabin, which can improve the accuracy of detection.
[0080] See also Figure 2 , Figure 2 A schematic flow chart of a method for detecting a living body in a vehicle provided by an embodiment of the present disclosure is shown. The method is executed by a vehicle.
[0081] Specifically, the in-vehicle liveness detection method includes:
[0082] S201, controlling the wireless communication system to transmit a wireless radio frequency signal to the interior of the vehicle to obtain radio frequency signal feedback data;
[0083] The specific process is as above and will not be repeated here.
[0084] S202, obtaining an image sample data set and a radio frequency signal feedback sample data set;
[0085] According to some embodiments, the image sample data set includes image sample data corresponding to at least one scene in the interior of a vehicle.
[0086] In some embodiments, the image sample data may be acquired by a camera system inside the vehicle. The image sample data and the RF signal feedback sample data at the same time correspond one to one.
[0087] It is easy to understand that when a vehicle performs in-vehicle living body detection, the vehicle can obtain an image sample data set and a radio frequency signal feedback sample data set.
[0088] S203, training the initial liveness detection model according to the image sample data set and the radio frequency signal feedback sample data set to obtain a target liveness detection model that meets the training conditions;
[0089] According to some embodiments, the initial living body detection model refers to an untrained neural network module.
[0090] In some embodiments, the training condition refers to the condition used by the vehicle to determine whether the initial liveness detection model has achieved the training target. The training condition may be, for example, that the detection accuracy during in-vehicle liveness detection reaches an accuracy threshold.
[0091] According to some embodiments, when the initial liveness detection model is trained according to the image sample data set and the radio frequency signal feedback sample data set to obtain the target liveness detection model that meets the training conditions, first, the initial liveness detection model can be trained according to the image sample data set and the radio frequency signal feedback sample data set to obtain the trained liveness detection model. Then, the trained initial liveness detection model can be tested using the radio frequency signal feedback test data set to obtain the detection accuracy corresponding to the trained liveness detection model. Secondly, if the detection accuracy is lower than the accuracy threshold, the trained liveness detection model is used as the initial liveness detection model for retraining until the detection accuracy is not lower than the accuracy threshold, and the target liveness detection model that meets the training conditions is obtained.
[0092] In some embodiments, the RF signal feedback test data set includes RF signal feedback test data corresponding to at least one scenario in the interior of the vehicle. The RF signal feedback test data refers to feedback data for testing obtained after transmitting a wireless RF signal to the interior of the vehicle.
[0093] In some embodiments, when the trained initial liveness detection model is tested using the radio frequency signal feedback test data set, the radio frequency signal feedback test data set can be input into the trained initial liveness detection model to obtain a liveness detection test result set. Thus, the detection accuracy corresponding to the trained liveness detection model can be determined based on the accuracy of the liveness detection test results in the liveness detection test result set.
[0094] For example, there are 100 liveness detection test results in the liveness detection test result set, but only 90 of them are correct. Then, the detection accuracy of the trained liveness detection model is 90%.
[0095] In some embodiments, the accuracy threshold does not specifically refer to a fixed value. The accuracy threshold can be adjusted according to the actual application scenario. The accuracy threshold can be, for example, 95%, or 90%.
[0096] According to some embodiments, when the initial liveness detection model is trained based on the image sample data set and the radio frequency signal feedback sample data set to obtain the trained liveness detection model, first, at least one image sample data in the image sample data set and at least one radio frequency signal feedback sample data in the radio frequency signal feedback sample data set can be labeled and classified to obtain the labeled and classified image sample data set and the labeled and classified radio frequency signal feedback sample data set. Then, the initial liveness detection model can be trained based on the labeled and classified image sample data set and the labeled and classified radio frequency signal feedback sample data set to obtain the trained liveness detection model.
[0097] In some embodiments, labeling and classification includes data labeling and classification. Data labeling is used to identify raw data (pictures, text files, videos, etc.) and add one or more meaningful information tags to provide context so that the model can learn from it. Data classification refers to grouping data with certain common attributes or characteristics together and distinguishing the data by the attributes or characteristics of its category.
[0098] In some embodiments, when the initial living body detection model is trained according to the labeled and classified image sample data set and the labeled and classified radio frequency signal feedback sample data set, at least one of the following training methods may be used:
[0099] Transfer learning training;
[0100] supervised learning training;
[0101] Distillation learning training.
[0102] In some embodiments, transfer learning training uses existing knowledge to learn new knowledge. Its core is to find the similarities between existing knowledge and new knowledge, and achieve the purpose of transfer learning through the transfer of this similarity. Supervised learning training refers to the process of adjusting the parameters of the classifier using a set of samples of known categories to achieve the required performance. It is also called supervised training or teacher learning. It is a machine learning task to infer a function from labeled training data. Distillation learning training is knowledge distillation, which transfers the knowledge of the teacher network to the student network so that the performance of the student network is the same as that of the teacher network.
[0103] For example, when the initial liveness detection model is trained according to the labeled and classified image sample data set and the labeled and classified RF signal feedback sample data set to obtain the trained liveness detection model, first, the initial liveness detection model can be trained according to the labeled and classified image sample data set to obtain the image-trained liveness detection model. Then, the image-trained liveness detection model can be trained according to the labeled and classified RF signal feedback sample data set to obtain the trained liveness detection model. Thus, the image-trained liveness detection model can be used to strengthen or supervise the process of training with labeled and classified RF signals, that is, the first image training is the teacher, the second RF signal training is the student, and the second volume is to use the teacher to guide the student's training. Therefore, the detection accuracy and generalization ability of the trained liveness detection model can be improved.
[0104] It is easy to understand that when the vehicle obtains an image sample data set and an RF signal feedback sample data set, the vehicle can train the initial liveness detection model based on the image sample data set and the RF signal feedback sample data set to obtain a target liveness detection model that meets the training conditions.
[0105] S204, controlling the target liveness detection model to perform liveness detection according to the radio frequency signal feedback data to obtain a liveness detection result.
[0106] Take a scenario as an example. Figure 3 FIG. 1 is a flow chart of a method for detecting a living body in a vehicle provided by an embodiment of the present disclosure. Figure 3 As shown, first, the vehicle can obtain a large amount of in-cabin video image data collection under multiple scenarios and a large amount of in-cabin RF signal data collection under multiple scenarios. Then, the vehicle can perform data labeling and classification on these collected data for the same scenario. Secondly, the vehicle can use the video image data after data labeling and classification to perform the first network model training, and use this to supervise the training process of the second network model training using the RF signal data after data labeling and classification. Finally, determine whether the trained liveness detection model meets the training conditions. If not, obtain the same scene data again and continue to train the model until the trained liveness detection model meets the training conditions to obtain an accurate prediction model, that is, a target liveness detection model. Finally, the vehicle can input the cabin RF signal detection data, that is, the RF signal feedback data, into the accurate prediction model to obtain the liveness detection result.
[0107] In summary, the method provided by the embodiment of the present disclosure can, first of all, control the wireless communication system to transmit wireless radio frequency signals to the interior of the vehicle to obtain radio frequency signal feedback data; therefore, by reusing the original wireless communication system in the vehicle for liveness detection, there is no need to set up a complete set of hardware systems separately, which can reduce the cost of in-vehicle liveness detection. At the same time, the radio frequency signal has strong penetration and diffraction capabilities and will not be blocked by obstacles, which can improve the accuracy of detection. Then, obtain an image sample data set and a radio frequency signal feedback sample data set; train the initial liveness detection model according to the image sample data set and the radio frequency signal feedback sample data set to obtain a target liveness detection model that meets the training conditions; control the target liveness detection model to perform liveness detection according to the radio frequency signal feedback data to obtain a liveness detection result; therefore, by combining artificial intelligence network training with radio frequency signals to realize liveness detection, extremely high technical, performance and cost benefits can be achieved; at the same time, through continuous big data training, the accuracy of detection can be continuously improved to cope with complex situations in the actual use environment.
[0108] The following are embodiments of the device disclosed herein, which can be used to execute the method embodiments disclosed herein. For details not disclosed in the device embodiments disclosed herein, please refer to the method embodiments disclosed herein.
[0109] See also Figure 4 , which shows a schematic diagram of the structure of an in-vehicle liveness detection device provided by an exemplary embodiment of the present disclosure. The in-vehicle liveness detection device can be implemented as all or part of the device through software, hardware, or a combination of both. The in-vehicle liveness detection device 400 includes a signal transceiver unit 401 and a liveness detection unit 402, wherein:
[0110] The signal transceiver unit 401 is used to control the wireless communication system to transmit a wireless radio frequency signal to the interior of the vehicle to obtain radio frequency signal feedback data;
[0111] The liveness detection unit 402 is used to control the target liveness detection model to perform liveness detection according to the RF signal feedback data to obtain a liveness detection result, wherein the target liveness detection model is trained according to the RF signal feedback sample data set, and the RF signal feedback sample data set includes RF signal feedback sample data corresponding to at least one scene inside the vehicle.
[0112] According to some embodiments, the in-vehicle liveness detection device 400 further includes a model training unit 403, which is used to, before controlling the target liveness detection model to perform liveness detection according to the radio frequency signal feedback data, specifically:
[0113] Acquire an image sample data set and a radio frequency signal feedback sample data set, wherein the image sample data set includes image sample data corresponding to at least one scene inside the vehicle, and the image sample data and the radio frequency signal feedback sample data correspond one to one;
[0114] The initial liveness detection model is trained according to the image sample data set and the radio frequency signal feedback sample data set to obtain a target liveness detection model that meets the training conditions.
[0115] According to some embodiments, the model training unit 403 is used to train the initial liveness detection model according to the image sample data set and the radio frequency signal feedback sample data set, and when the target liveness detection model that meets the training conditions is obtained, it is specifically used to:
[0116] Training the initial liveness detection model according to the image sample data set and the radio frequency signal feedback sample data set to obtain a trained liveness detection model;
[0117] The trained initial liveness detection model is tested using a radio frequency signal feedback test data set to obtain a detection accuracy corresponding to the trained liveness detection model, wherein the radio frequency signal feedback test data set includes radio frequency signal feedback test data corresponding to at least one scenario inside a vehicle;
[0118] If the detection accuracy is lower than the accuracy threshold, the trained liveness detection model is used as the initial liveness detection model for retraining until the detection accuracy is not lower than the accuracy threshold, thereby obtaining a target liveness detection model that meets the training conditions.
[0119] According to some embodiments, the model training unit 403 is used to train the initial liveness detection model according to the image sample data set and the radio frequency signal feedback sample data set, and when the trained liveness detection model is obtained, it is specifically used to:
[0120] Marking and classifying at least one image sample data in the image sample data set and at least one radio frequency signal feedback sample data in the radio frequency signal feedback sample data set to obtain a marked and classified image sample data set and a marked and classified radio frequency signal feedback sample data set;
[0121] The initial liveness detection model is trained according to the labeled and classified image sample data set and the labeled and classified radio frequency signal feedback sample data set to obtain a trained liveness detection model.
[0122] According to some embodiments, the model training unit 403 is used to train the initial living body detection model according to the labeled and classified image sample data set and the labeled and classified RF signal feedback sample data set, using at least one of the following training methods:
[0123] Transfer learning training;
[0124] supervised learning training;
[0125] Distillation learning training.
[0126] According to some embodiments, the model training unit 403 is used to train the initial liveness detection model according to the labeled and classified image sample data set and the labeled and classified radio frequency signal feedback sample data set, and when the trained liveness detection model is obtained, it is specifically used to:
[0127] The initial liveness detection model is trained according to the labeled and classified image sample data set to obtain a liveness detection model after image training;
[0128] The liveness detection model trained by the image is trained according to the radio frequency signal feedback sample data set after labeling and classification, so as to obtain the trained liveness detection model.
[0129] It should be noted that: Figure 4 As shown, the modules that must be included in the in-vehicle liveness detection device 400 are indicated by solid-line boxes, such as the signal transceiver unit 401 and the liveness detection unit 402; the modules that may or may not be included in the in-vehicle liveness detection device 400 are indicated by dotted-line boxes, such as the model training unit 403.
[0130] Regarding the device in the above embodiment, the specific manner in which each module performs operations has been described in detail in the embodiment of the method, and will not be elaborated here.
[0131] In summary, the device provided by the embodiment of the present disclosure controls the wireless communication system to transmit wireless radio frequency signals to the interior of the vehicle through a signal transceiver unit to obtain radio frequency signal feedback data; the liveness detection unit controls the target liveness detection model to perform liveness detection according to the radio frequency signal feedback data to obtain a liveness detection result, wherein the target liveness detection model is trained based on a set of radio frequency signal feedback sample data, and the set of radio frequency signal feedback sample data includes radio frequency signal feedback sample data corresponding to at least one scene inside the vehicle. Therefore, by reusing the original wireless communication system in the vehicle for liveness detection, there is no need to set up a complete hardware system separately, which can reduce the cost of performing liveness detection in the vehicle. At the same time, the radio frequency signal has strong penetration and diffraction capabilities and will not be blocked by obstacles, which can improve the accuracy of detection.
[0132] According to an embodiment of the present disclosure, the present disclosure also provides a vehicle, a readable storage medium, and a computer program product.
[0133] Figure 5A schematic block diagram of an example vehicle 500 that may be used to implement embodiments of the present disclosure is shown. The components shown herein, their connections and relationships, and their functions are provided as examples only, and are not intended to limit implementations of the present disclosure described and / or claimed herein.
[0134] like Figure 5 As shown, the vehicle 500 includes a computing unit 501, which can perform various appropriate actions and processes according to a computer program stored in a read-only memory (ROM) 502 or a computer program loaded from a storage unit 508 into a random access memory (RAM) 503. In the RAM 503, various programs and data required for the operation of the vehicle 500 can also be stored. The computing unit 501, the ROM 502, and the RAM 503 are connected to each other via a bus 504. An input / output (I / O) interface 505 is also connected to the bus 504.
[0135] A number of components in the vehicle 500 are connected to the I / O interface 505, including: an input unit 506, such as a keyboard, a mouse, etc.; an output unit 507, such as various types of displays, speakers, etc.; a storage unit 508, such as a disk, an optical disk, etc.; and a communication unit 509, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 509 allows the vehicle 500 to exchange information / data with other devices through a computer network such as the Internet and / or various telecommunication networks.
[0136] The computing unit 501 may be a variety of general and / or special processing components with processing and computing capabilities. Some examples of the computing unit 501 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various dedicated artificial intelligence (AI) computing chips, various computing units running machine learning model algorithms, digital signal processors (DSPs), and any appropriate processors, controllers, microcontrollers, etc. The computing unit 501 performs the various methods and processes described above, such as the in-vehicle liveness detection method. For example, in some embodiments, the in-vehicle liveness detection method may be implemented as a computer software program, which is tangibly contained in a machine-readable medium, such as a storage unit 508. In some embodiments, part or all of the computer program may be loaded and / or installed on the vehicle 500 via the ROM 502 and / or the communication unit 509. When the computer program is loaded into the RAM 503 and executed by the computing unit 501, one or more steps of the in-vehicle liveness detection method described above may be performed. Alternatively, in other embodiments, the computing unit 501 may be configured to perform the in-vehicle liveness detection method in any other appropriate manner (e.g., by means of firmware).
[0137] Various implementations of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chips (SOCs), load programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various implementations can include: being implemented in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which can be a special purpose or general purpose programmable processor that can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the storage system, the at least one input device, and the at least one output device.
[0138] The program code for implementing the method of the present disclosure may be written in any combination of one or more programming languages. These program codes may be provided to a processor or controller of a general-purpose computer, a special-purpose computer, or other programmable data processing device, so that the program code, when executed by the processor or controller, causes the functions / operations specified in the flow chart and / or block diagram to be implemented. The program code may be executed entirely on the machine, partially on the machine, partially on the machine and partially on a remote machine as a stand-alone software package, or entirely on a remote machine or vehicle.
[0139] In the context of the present disclosure, a machine-readable medium may be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, device, or equipment. A machine-readable medium may be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium may include, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, device, or equipment, or any suitable combination of the foregoing. A more specific example of a machine-readable storage medium may include an electrical connection based on one or more lines, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.
[0140] To provide interaction with a user, the systems and techniques described herein can be implemented on a computer having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and pointing device (e.g., a mouse or trackball) through which the user can provide input to the computer. Other types of devices can also be used to provide interaction with the user; for example, the feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including acoustic input, voice input, or tactile input).
[0141] The systems and techniques described herein may be implemented in a computing system that includes backend components (e.g., as a data vehicle), or a computing system that includes middleware components (e.g., an application vehicle), or a computing system that includes frontend components (e.g., a user computer with a graphical user interface or a web browser through which a user can interact with implementations of the systems and techniques described herein), or a computing system that includes any combination of such backend components, middleware components, or frontend components. The components of the system may be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: a local area network (LAN), a wide area network (WAN), the Internet, and a blockchain network.
[0142] The computer system may include a client and a vehicle. The client and the vehicle are generally far away from each other and usually interact through a communication network. The relationship between the client and the vehicle is generated by computer programs running on corresponding computers and having a client-vehicle relationship with each other. The vehicle can be a cloud vehicle, also known as a cloud computing vehicle or a cloud host, which is a host product in the cloud computing service system to solve the defects of difficult management and weak business scalability in traditional physical hosts and VPS services ("Virtual Private Server", or "VPS" for short). The vehicle can also be a vehicle in a distributed system, or a vehicle combined with a blockchain.
[0143] It should be understood that the various forms of processes shown above can be used to reorder, add or delete steps. For example, the steps recorded in this disclosure can be executed in parallel, sequentially or in different orders, as long as the desired results of the technical solutions disclosed in this disclosure can be achieved, and this document does not limit this.
[0144] The above specific implementations do not constitute a limitation on the protection scope of the present disclosure. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions can be made according to design requirements and other factors. Any modification, equivalent substitution and improvement made within the spirit and principle of the present disclosure shall be included in the protection scope of the present disclosure.
Claims
1. A method for detecting a living body in a vehicle, characterized in that: Applied to vehicles, including: Control the wireless communication system to transmit a wireless radio frequency signal to the interior of the vehicle to obtain radio frequency signal feedback data; The target liveness detection model is controlled to perform liveness detection according to the RF signal feedback data to obtain a liveness detection result, wherein the target liveness detection model is trained according to a set of RF signal feedback sample data, and the set of RF signal feedback sample data includes RF signal feedback sample data corresponding to at least one scene inside the vehicle.
2. The method according to claim 1, characterized in that: Before the control target liveness detection model performs liveness detection according to the radio frequency signal feedback data, the method further includes: Acquire an image sample data set and a radio frequency signal feedback sample data set, wherein the image sample data set includes image sample data corresponding to at least one scene inside the vehicle, and the image sample data and the radio frequency signal feedback sample data correspond one to one; The initial liveness detection model is trained according to the image sample data set and the radio frequency signal feedback sample data set to obtain a target liveness detection model that meets the training conditions.
3. The method according to claim 2, characterized in that The initial liveness detection model is trained according to the image sample data set and the radio frequency signal feedback sample data set to obtain a target liveness detection model that meets the training conditions, including: Training an initial liveness detection model according to the image sample data set and the radio frequency signal feedback sample data set to obtain a trained liveness detection model; Using a radio frequency signal feedback test data set to test the trained initial liveness detection model to obtain a detection accuracy corresponding to the trained liveness detection model, wherein the radio frequency signal feedback test data set includes radio frequency signal feedback test data corresponding to at least one scene inside the vehicle; If the detection accuracy is lower than the accuracy threshold, the trained liveness detection model is retrained as the initial liveness detection model until the detection accuracy is not lower than the accuracy threshold, thereby obtaining a target liveness detection model that meets the training conditions.
4. The method according to claim 3, characterized in that The training of the initial liveness detection model according to the image sample data set and the radio frequency signal feedback sample data set to obtain the trained liveness detection model includes: marking and classifying at least one image sample data in the image sample data set and at least one radio frequency signal feedback sample data in the radio frequency signal feedback sample data set to obtain a marked and classified image sample data set and a marked and classified radio frequency signal feedback sample data set; The initial living body detection model is trained according to the labeled and classified image sample data set and the labeled and classified radio frequency signal feedback sample data set to obtain a trained living body detection model.
5. The method according to claim 4, characterized in that When the initial living body detection model is trained according to the labeled and classified image sample data set and the labeled and classified radio frequency signal feedback sample data set, at least one of the following training methods is adopted: Transfer learning training; supervised learning training; Distillation learning training.
6. The method according to claim 4, characterized in that The initial living body detection model is trained according to the labeled and classified image sample data set and the labeled and classified radio frequency signal feedback sample data set to obtain the trained living body detection model, including: Training the initial liveness detection model according to the labeled and classified image sample data set to obtain a liveness detection model after image training; The liveness detection model trained with the image is trained according to the radio frequency signal feedback sample data set after the label classification to obtain the trained liveness detection model.
7. A liveness detection device in a vehicle, characterized in that: Applied to vehicles, including: A signal transceiver unit, used to control the wireless communication system to transmit a wireless radio frequency signal to the interior of the vehicle and obtain radio frequency signal feedback data; A liveness detection unit is used to control a target liveness detection model to perform liveness detection according to the RF signal feedback data to obtain a liveness detection result, wherein the target liveness detection model is trained based on a set of RF signal feedback sample data, and the set of RF signal feedback sample data includes RF signal feedback sample data corresponding to at least one scene inside the vehicle.
8. The device according to claim 7, characterized in that The device also includes a model training unit, which is used for, before the control target living body detection model performs living body detection according to the radio frequency signal feedback data, specifically for: Acquire an image sample data set and a radio frequency signal feedback sample data set, wherein the image sample data set includes image sample data corresponding to at least one scene inside the vehicle, and the image sample data and the radio frequency signal feedback sample data correspond one to one; The initial liveness detection model is trained according to the image sample data set and the radio frequency signal feedback sample data set to obtain a target liveness detection model that meets the training conditions.
9. The device according to claim 8, characterized in that The model training unit is used to train the initial liveness detection model according to the image sample data set and the radio frequency signal feedback sample data set to obtain a target liveness detection model that meets the training conditions, specifically for: Training an initial liveness detection model according to the image sample data set and the radio frequency signal feedback sample data set to obtain a trained liveness detection model; Using a radio frequency signal feedback test data set to test the trained initial liveness detection model to obtain a detection accuracy corresponding to the trained liveness detection model, wherein the radio frequency signal feedback test data set includes radio frequency signal feedback test data corresponding to at least one scene inside the vehicle; If the detection accuracy is lower than the accuracy threshold, the trained liveness detection model is retrained as the initial liveness detection model until the detection accuracy is not lower than the accuracy threshold, thereby obtaining a target liveness detection model that meets the training conditions.
10. The device according to claim 9, characterized in that The model training unit is used to train the initial liveness detection model according to the image sample data set and the radio frequency signal feedback sample data set to obtain the trained liveness detection model, specifically for: marking and classifying at least one image sample data in the image sample data set and at least one radio frequency signal feedback sample data in the radio frequency signal feedback sample data set to obtain a marked and classified image sample data set and a marked and classified radio frequency signal feedback sample data set; The initial living body detection model is trained according to the labeled and classified image sample data set and the labeled and classified radio frequency signal feedback sample data set to obtain a trained living body detection model.
11. The device according to claim 10, characterized in that The model training unit is used to train the initial living body detection model according to the labeled and classified image sample data set and the labeled and classified radio frequency signal feedback sample data set, using at least one of the following training methods: Transfer learning training; supervised learning training; Distillation learning training.
12. The device according to claim 10, characterized in that The model training unit is used to train the initial liveness detection model according to the labeled and classified image sample data set and the labeled and classified radio frequency signal feedback sample data set, and when the trained liveness detection model is obtained, it is specifically used to: Training the initial liveness detection model according to the labeled and classified image sample data set to obtain a liveness detection model after image training; The liveness detection model trained with the image is trained according to the radio frequency signal feedback sample data set after the label classification to obtain the trained liveness detection model.
13. A vehicle, characterized in that: include: processor; a memory for storing instructions executable by the processor; The processor is configured to execute the instructions to implement the in-vehicle living body detection method as described in any one of claims 1 to 6.
14. A storage medium, when instructions in the storage medium are executed by a processor of a vehicle, the vehicle is enabled to execute the in-vehicle living body detection method according to any one of claims 1 to 6.