Biological information conversion method and device, intelligent vehicle, and storage medium

Abstract

The application discloses a kind of biological information transformation method, device, intelligent automobile and storage medium, the method includes: obtaining the original biological information of user, and based on the original biological information, and preset biological information generation model, generate the transformation biological information corresponding to the original biological information;Based on the transformation biological information, the preset service model under target service is trained, and based on the service model that has been trained, the transformation biological information is handled, and the first service data corresponding to the transformation biological information is obtained;If the error between the first service data and the second service data of the original biological information under the target service is less than preset error, then the transformation biological information is used as the alternative information of the original biological information.The above scheme can effectively protect the personal data of user.

Description

Technical Field

[0001] This invention relates to the field of vehicle technology, and in particular to a bio-information conversion method, device, intelligent vehicle, and storage medium. Background Technology

[0002] With the continuous development of science and technology, the personalized functions of products have received widespread attention. In order to provide users with more convenient and seamless services, more and more personalized functions are implemented using users' biometric information (such as facial information, iris information, etc.). However, biometric information is a type of personal data and involves user privacy. Therefore, how to protect users' biometric information is an urgent problem to be solved. Summary of the Invention

[0003] This invention provides a bio-information transformation method, device, intelligent vehicle, and storage medium to effectively protect users' real bio-information.

[0004] In a first aspect, embodiments of the present invention provide a bioinformatics transformation method, comprising:

[0005] Obtain the user's original biological information, and based on the original biological information and a preset biological information generation model, generate transformed biological information corresponding to the original biological information;

[0006] Based on the transformed biological information, a preset business model under the target business is trained, and the transformed biological information is processed based on the trained business model to obtain the first business data corresponding to the transformed biological information.

[0007] If the error between the first business data and the original biological information in the second business data under the target business is less than a preset error, then the transformed biological information will be used as the substitute information for the original biological information.

[0008] In some implementations, generating transformed biological information corresponding to the original biological information based on the original biological information and a preset biological information generation model includes:

[0009] Determine whether the original training sample set of the preset bioinformatics generation model contains N first bioinformatics, wherein the similarity between the features corresponding to the N first bioinformatics and the features corresponding to the original bioinformatics is greater than or equal to a preset similarity, and N is a positive integer;

[0010] If the original training sample set does not contain the N first biological information, based on the original biological information, M second biological information are determined for adjusting the preset biological information generation model, wherein the similarity between the features corresponding to the M second biological information and the features corresponding to the original biological information is greater than or equal to the preset similarity, and M is a positive integer;

[0011] The preset biological information is updated based on the M second biological information, and the transformed biological information is output.

[0012] In some implementations, training a preset business model for the target business based on the transformed biometric information includes:

[0013] Based on the transformed biological information, a first initial feature extraction model is trained, and based on the trained first feature extraction model, feature data of the transformed biological information is extracted.

[0014] The preset business model is trained based on the feature data of the transformed biological information.

[0015] The data output range of the first feature extraction model is the same as that of the second feature extraction model, and the second feature extraction model is a model used to extract features from the original biological information.

[0016] In some implementations, before using the transformed biological information as a substitute for the original biological information if the error between the first service data and the original biological information in the second service data under the target service is less than a preset error, the method further includes:

[0017] Based on the second feature extraction model, feature data of the original biological information are extracted;

[0018] Based on the preset business model, the feature data of the original biological information are processed to obtain the second business data corresponding to the original biological information.

[0019] In some implementations, after obtaining the first business data corresponding to the transformed biometric information, the method further includes:

[0020] If the error between the first business data and the second business data is greater than or equal to the preset error, the steps of generating new transformed biological information through the preset biological information generation model and training the preset business model based on the new transformed biological information to obtain new business data corresponding to the new transformed biological information are repeated until the error between the obtained business data and the second business data is less than the preset error.

[0021] In some implementations, obtaining the user's original biometric information includes:

[0022] During the validity period of the user's original biometric information, the user's original biometric information is obtained.

[0023] In some implementations, the original biometric information includes the user's real image information, and the preset biometric information generation model includes a generative adversarial network model.

[0024] Secondly, embodiments of the present invention provide a bioinformatics conversion device, comprising:

[0025] The first processing module is used to acquire the user's original biological information and, based on the original biological information and a preset biological information generation model, generate transformed biological information corresponding to the original biological information.

[0026] The business data generation module is used to train a preset business model under the target business based on the transformed biological information, and to process the transformed biological information based on the trained business model to obtain the first business data corresponding to the transformed biological information.

[0027] The second processing module is configured to use the transformed biological information as a substitute for the original biological information if the error between the first business data and the original biological information in the second business data under the target business is less than a preset error.

[0028] Thirdly, the present invention provides an intelligent vehicle, including a memory and one or more programs, wherein the one or more programs are stored in the memory and configured to be executed by one or more processors, comprising operation instructions for performing any of the methods provided in the first aspect.

[0029] Fourthly, embodiments of the present invention provide a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of any of the methods provided in the first aspect.

[0030] The above-described one or more technical solutions in the embodiments of this application have at least the following technical effects:

[0031] The biometric information transformation method provided in this specification generates transformed biometric information corresponding to the original biometric information based on the user's original biometric information and a preset biometric information generation model. Based on the transformed biometric information, a preset business model under the target business is trained, and the transformed biometric information is processed based on the trained business model to obtain first business data corresponding to the transformed biometric information. If the error between the first business data and the original biometric information under the target business is less than a preset error, the transformed biometric information is used as the original biometric information. In this solution, by transforming the original biometric information, transformed biometric information similar to the original biometric information under the target business is obtained. Since the transformed biometric information can meet the business requirements of the target business, it can replace the real biometric information, thereby effectively protecting the user's personal data. Attached Figure Description

[0032] Figure 1 A flowchart of a bioinformatics transformation method provided in the embodiments of this specification;

[0033] Figure 2 This specification provides a schematic diagram of a processing flow for raw biological information, as illustrated in an embodiment.

[0034] Figure 3 This specification provides a schematic diagram of a processing flow for transforming biological information, as illustrated in an embodiment of the invention.

[0035] Figure 4 A schematic diagram of a bio-information conversion device provided in the embodiments of this specification;

[0036] Figure 5 This is a schematic diagram of an intelligent vehicle provided as an embodiment of this specification. Detailed Implementation

[0037] The overall technical solution of this application embodiment is as follows: Obtain the user's original biological information, and based on the original biological information and a preset biological information generation model, generate transformed biological information corresponding to the original biological information; based on the transformed biological information, train a preset business model under the target business, and process the transformed biological information based on the trained business model to obtain first business data corresponding to the transformed biological information; if the error between the first business data and the original biological information under the target business is less than a preset error, then use the transformed biological information as a substitute for the original biological information.

[0038] The solution provided in the embodiments of this specification transforms the original biometric information to obtain transformed biometric information that is similar to the original biometric information under the target business. Since the transformed biometric information can meet the business requirements of the target business, the transformed biometric information can replace the real biometric information, thereby effectively protecting the user's personal data.

[0039] To better understand the above technical solutions, the technical solutions of the embodiments of this specification will be described in detail below with reference to the accompanying drawings and specific embodiments. It should be understood that the embodiments of this specification and the specific features in the embodiments are detailed descriptions of the technical solutions of the embodiments of this specification, rather than limitations on the technical solutions of this specification. Unless otherwise specified, the embodiments of this specification and the technical features in the embodiments can be combined with each other.

[0040] First, it should be clarified that the term "and / or" in this article is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, or B existing alone. Additionally, the character " / " in this article generally indicates that the preceding and following related objects have an "or" relationship.

[0041] like Figure 1 The diagram shown is a flowchart of a bioinformatics transformation method provided in an embodiment of this specification. The method includes the following steps:

[0042] Step S101: Obtain the user's original biological information, and based on the original biological information and a preset biological information generation model, generate transformed biological information corresponding to the original biological information;

[0043] Step S102: Based on the transformed biological information, train a preset business model under the target business, and process the transformed biological information based on the trained business model to obtain the first business data corresponding to the transformed biological information;

[0044] Step S103: If the error between the first service data and the original biological information in the second service data under the target service is less than a preset error, then the transformed biological information is used as the replacement information for the original biological information.

[0045] The methods described in the embodiments of this specification can be applied to terminal devices, such as user mobile phones, computers, etc., or to user vehicles. These terminal devices can provide personalized services based on the user's biometric information. The methods described in the embodiments of this specification can also be applied to servers, such as servers corresponding to vehicles, cloud servers corresponding to mobile phones, etc., and can also be applied to systems composed of terminal devices and servers; no limitations are set herein.

[0046] In step S101, the user's original biometric information refers to the user's real biometric information, which may include, but is not limited to, the user's image information, iris information, etc. To better provide personalized services, many functions utilize this original biometric information, achieving personalized services without the user's awareness and improving the user experience. For example, in the vehicle field, a preset recognition model is used to recognize the user's face, and personalized parameters are determined based on the recognition results to adjust the vehicle's driving data and cabin data, thereby providing the user with a better driving experience.

[0047] However, users' original biometric information involves user privacy and is considered sensitive data. To prevent privacy leaks, users typically authorize the validity period of their original biometric information. For example, in vehicle services, users can authorize vehicle functions to use facial image information, iris information, and other original biometric information for a validity period of 1 month, 3 months, or 1 year. Once the validity period expires, the original biometric information can no longer be used.

[0048] In the embodiments described in this specification, in order to continue providing accurate personalized services to users after the expiration date, the original biometric information can be transformed to obtain transformed biometric information that can replace the original biometric information. It should be noted that the transformed biometric information and the original biometric information are completely different, but some features of the transformed biometric information are similar to some features of the original biometric information. Taking a face image as an example, the transformed biometric information can be a face image transformed from the face image. However, the original face image cannot be reversed from the transformed face image. Therefore, it can effectively protect the user's personal data while allowing the generated transformed biometric information to replace the original biometric information to complete business functions.

[0049] In the embodiments of this specification, the preset biological information generation model can be set according to actual needs. For example, the preset biological information generation model can be a diffusion model (DM), a generative adversarial network (GAN), etc., and is not limited here. Transformed biological information can be generated using the preset biological information generation model and the original biological information.

[0050] In some embodiments, the generation of transformed biological information can be achieved through the following steps: determining whether the original training sample set of the preset biological information generation model contains N first biological information, wherein the similarity between the features corresponding to the N first biological information and the features corresponding to the original biological information is greater than or equal to a preset similarity, and N is a positive integer; if the original training sample set does not contain the N first biological information, determining M second biological information for adjusting the preset biological information generation model based on the original biological information, wherein the similarity between the features corresponding to the M second biological information and the features corresponding to the original biological information is greater than or equal to the preset similarity, and M is a positive integer; updating the preset biological information based on the M second biological information, and outputting the transformed biological information.

[0051] Specifically, the preset biometrics generation model can be a pre-trained model. Taking StyleGAN2 as an example, StyleGAN2 can be trained based on a public database and can generate face images. When the original biometrics is the user's real face image, in order to ensure that StyleGAN2 can generate a transformed face image that approximates the user's real face image, it is necessary to first determine whether the original training sample set of StyleGAN2 (such as a public database) contains N first biometrics, where N is a positive integer.

[0052] It should be noted that the N first biological information contains some or all of the features of the original biological information. That is, the similarity between the features corresponding to the N first biological information and the features corresponding to the original biological information is greater than or equal to the preset similarity. The preset similarity can be set according to actual needs. For example, the preset similarity can be 85%, 90%, etc.

[0053] For example, if the user is a male with glasses and black hair, and the images in the original training sample set of StyleGAN2 do not contain features indicating wearing glasses and having black hair, then the StyleGAN2 model cannot generate a face image of a person wearing glasses. Consequently, the resulting transformed face image will lack the user's key features. To avoid this, in this embodiment, the ability of the StyleGAN2 model to generate biometric information similar to the original biometric information is determined by whether the original training sample set contains N pieces of first biometric information.

[0054] It should be understood that if N is 1, then the image needs to contain some or all of the features in the original biological information; if N is an integer greater than 1, then the features in these N images must contain some or all of the features in the original biological information. If the features of the original biological information include wearing glasses, black hair, and wearing a hat, and the preset similarity is 65%, then it is possible to detect whether there is an image in the original training sample set that contains at least two of the features of wearing glasses, black hair, and wearing a hat, or to statistically analyze the features in each image in the original training sample set. If the statistical results show that the original training set contains at least two of the features of wearing glasses, black hair, and wearing a hat, then it indicates that the original training sample set contains N pieces of first biological information.

[0055] If the original training sample set contains the above N first biological information, it indicates that the training sample set covers most or all of the features in the original biological information, and the preset biological information generation model can be used directly to generate transformed biological information.

[0056] If the original training sample set does not contain the aforementioned N first biological information items, it indicates that the existing preset biological information generation model cannot generate transformed biological information similar to the original biological information. In this case, based on the original biological information, M second biological information items can be determined, wherein the similarity between the features corresponding to the M second biological information items and the features corresponding to the original biological information items is greater than the preset similarity.

[0057] Continuing with the original biometric information example, which includes wearing glasses, having black hair, and wearing a hat, and assuming a similarity of 65%, for M pieces of second biometric information, it is sufficient to ensure that the features statistically analyzed from the M pieces of second biometric information cover any two of the above three features. For example, some second biometric information may be a face image of a person wearing glasses, some may be a face image of a person with black hair, some may be a face image of a person wearing a hat, and some may be a face image of a person with both black hair and glasses, etc.

[0058] By using M second biological information, the preset biological information generation model is fine-tuned. Taking the preset biological information generation model as StyleGAN2 as an example, the transformed biological information can be generated during the training process of StyleGAN2.

[0059] In step S102, in order to determine whether the transformed biological information can be used in place of the original biological information, in this embodiment of the specification, the preset business model under the target business is trained by transforming the biological information, and the first business data corresponding to the transformed biological information is obtained.

[0060] The target business can be a business that utilizes biometric information, such as providing personalized services to users based on biometric information. In some embodiments, the target business in the vehicle field may include identifying the vehicle owner based on biometric information, adjusting the vehicle seat based on biometric information, adjusting the ambient lighting in the vehicle based on biometric information, and predicting and navigating the user's current travel route based on biometric information.

[0061] The target business has a corresponding preset business model. Taking the target business as vehicle owner identification based on biometric information as an example, the preset business model can be a preset identification model. For example, if a vehicle has two owners, namely user 1 and user 2, then the preset business model can determine whether the biometric information corresponds to user 1, user 2, or other users based on the input biometric information.

[0062] It should be noted that within the validity period of the original biometric information, the preset business model for the target service can be trained and updated using the original biometric information to ensure that the preset business model can meet the user's personalized needs. However, after the validity period expires, the original biometric information can no longer be used. If the transformed biometric information generated by the preset biometric information generation model can replace the original biometric information, the transformed biometric information can be used to train and update the preset business model to continuously provide users with services under the target service.

[0063] In the embodiments of this specification, in order to determine whether the transformed biological information can replace the original biological information, a preset business model can be trained by transforming the biological information. After the training is completed, the transformed biological information is used as input data into the trained business model to output the first business data corresponding to the transformed biological information.

[0064] The preset business model can be a model trained using the original biological information within the validity period of the original biological information, or it can be a model trained using other biological information that can replace the original biological information. There are no restrictions on this.

[0065] Taking a face image as the biometric information and a pre-defined face recognition model as the preset business model as an example, the input data of the preset face recognition model is the face image, and the output data is the recognition result of the face image. The recognition result can be a user identifier. After obtaining the transformed biometric information, i.e., the transformed face image, the preset face recognition model is trained using the transformed face image. Based on the trained face recognition model, the user identifier corresponding to the transformed face image is obtained, and the identified user identifier is used as the first business data corresponding to the transformed face image.

[0066] In some implementations, training a preset business model based on transformed biological information can be achieved through the following steps: training a first initial feature extraction model based on the transformed biological information, and extracting feature data of the transformed biological information based on the trained first feature extraction model; training the preset business model based on the feature data of the transformed biological information; wherein the data output range of the first feature extraction model is the same as the data output range of the second feature extraction model, and the second feature extraction model is a model used to extract features from the original biological information.

[0067] Specifically, the initial feature extraction model can be set according to actual needs, such as a neural network model. Taking biological information as facial information, and the preset biological information generation model as the StyleGAN2 model as an example, the StyleGAN2 model can generate one or more transformed facial images. The initial feature extraction model is trained based on the generated biological information, and the feature data of each transformed facial image is obtained based on the trained initial feature extraction model.

[0068] It should be noted that, in order to determine whether the transformed biological information can replace the original biological information under the target business, the feature range corresponding to the transformed biological information needs to be set to be consistent with the feature range of the original biological information. In this way, the second business data of the original biological information under the target business can be used as a reference or evaluation basis to determine whether the first business data of the transformed biological information is similar to the second business data, thereby determining whether the transformed biological information can be used as a substitute for the original biological information.

[0069] In the specific implementation process, the output range of the first feature extraction model can be set to be the same as the output range of the second feature extraction model, wherein the second feature extraction model is a model used to extract features from the original biological information.

[0070] Furthermore, the feature data of the transformed biological information extracted by the first feature extraction model is used to train the preset business model, and the transformed biological information is processed based on the trained business model to obtain the first business data.

[0071] In step S103, the first business data and the second business data are compared. Since the second business data is the business data of the original biological information under the target business, if the error between the two is less than the preset error, it indicates that the performance of the transformed biological information under the target business is similar to that of the original biological information under the target business. That is, the transformed biological information and the original biological information can obtain nearly similar indicator data under the target business. Therefore, under the target business, the transformed biological information can be used as a substitute for the original biological information. It should be noted that the preset error can be set according to actual needs and is not limited here.

[0072] In the specific implementation process, the second business data can be obtained through the following steps: extracting the feature data of the original biological information based on the second feature extraction model; processing the feature data of the original biological information based on the preset business model to obtain the second business data corresponding to the original biological information.

[0073] Specifically, the second feature extraction model can be set according to actual needs, such as a neural network model. The structure of the second feature extraction model can be the same as or different from that of the first feature extraction model; there are no restrictions here. The second feature extraction model can be trained based on the original biological information. Based on the trained second feature extraction model, feature data of the original biological information can be extracted.

[0074] Furthermore, the feature data of the original biological information are processed through a pre-defined business model to obtain second business data. This second business data can serve as reference data for evaluating whether the transformed biological information can be used as a substitute for the original biological information.

[0075] In this embodiment of the specification, if the error between the first business data and the second business data is greater than or equal to the preset error, the steps of generating new transformed biological information through the preset biological information generation model and training the preset business model based on the new transformed biological information to obtain new business data corresponding to the new transformed biological information are repeated until the error between the obtained business data and the second business data is less than the preset error.

[0076] Specifically, if the error between the first and second business data is greater than or equal to a preset error, it indicates that the transformed biological information cannot replace the original biological information. In this case, it is necessary to regenerate new transformed biological information using a preset biological information generation model. Taking the StyleGAN2 model as an example, to generate new transformed biological information, different noise values ​​can be input. Further, after generating the new transformed biological information, the preset business model is retrained based on it to obtain a retrained business model. The new transformed biological information is then used as input data into the retrained business model to obtain new business data corresponding to the new transformed biological information. The error between the new business data and the second business data is compared again. If the error is less than the preset error, the new transformed biological information is used as a substitute for the original biological information. If the error is greater than or equal to the preset error, the above process is repeated.

[0077] Through the above methods, transformed biometric information that can replace the original biometric information for the target business can be obtained. It should be understood that within the validity period of the user's original biometric information, the original biometric information can be acquired and used to train or update the business model for the target business. If the original biometric information can no longer be used after the validity period expires, transformed biometric information that can replace the original biometric information can be generated using the method provided in the embodiments of this specification. This transformed biometric information is then saved. Therefore, even after the validity period expires, the saved transformed biometric information can still be used to update the business model for the target business and develop new functions, meeting the needs of the target business without infringing on the user's privacy data, thereby improving the user experience.

[0078] To better understand the methods provided in the embodiments of this specification, the following explanation uses a face image as the biometric information, the StyleGAN2 model as the preset biometric information generation model, and the vehicle owner identification service of the target vehicle as an example.

[0079] In practice, the target vehicle can have two owners: Owner 1 is a male owner, and Owner 2 is a female owner. During the validity period of the owner's authorization to use the original biometric information, real facial images of Owner 1 and Owner 2 (i.e., the original biometric information) can be collected and used. Based on these real facial images, a feature extraction function F is trained. raw (.), F raw (.) corresponds to the second feature extraction model described above. For ease of explanation, assume F raw The output of (.) (i.e., the extracted feature data) is an integer between 0 and 255. For simplicity, as shown below... Figure 2 As shown, assume that the target service needs to perform Figure 2 recognition on 5 face images in , and the preset service model under this target service is E raw (.), and E raw (.) outputs the recognition result of the image.

[0080] Please refer to Figure 2 , the image corresponding to image number 1 is the face image of vehicle owner 1, the image corresponding to image number 2 is the face image of vehicle owner 2, and images 3 to 5 are face images of non-vehicle owners, and each face image is processed through F raw (.), as Figure 2 shown, the output values of F raw (.) corresponding to images 1 to 5 are 12, 130, 69, 193, and 37 respectively. Further, E raw (.) processes the output value of F raw (.), wherein if the output value f of F raw (.) satisfies 0<f<16, the image is determined to be user 1, that is, the output value of E raw (.) is 1; if the output value f of F raw (.) satisfies 128<f<144, the image is determined to be user 2, that is, the output value of E raw (.) is 2; if the output value f of F raw (.) falls in other ranges, the image is determined to be a non-vehicle owner user, that is, the output value of E raw (.) is 0.

[0081] Further, the StyleGAN2 model can be fine-tuned using the real face images of the vehicle owners, or the StyleGAN2 model can be directly called to generate transformed biometric information, that is, generate transformed face images corresponding to vehicle owner 1 and vehicle owner 2. Based on the transformed face images corresponding to vehicle owner 1 and vehicle owner 2, the feature extraction function F is trained gen (.), F gen (.) corresponds to the aforementioned first feature extraction model, and F gen (.) has the same output range as F raw (.), both being integers from 0 to 255.

[0082] Please refer to Figure 3 , in order to determine whether the transformed face images can replace the real face information to meet the requirements of the target service, perform recognition on Figure 3 the 5 face images in , wherein the image corresponding to image number 1 is the generated transformed face image of vehicle owner 1, the image corresponding to image number 2 is the transformed face image of vehicle owner 2, and images 3 to 5 are face images of non-vehicle owners, and F is used gen(.) processing each face image, for example Figure 3 as shown, the output values of F corresponding to Images 1 to 5 are 68, 149, 37, 251, and 15 respectively. Further, through E gen (.) processing the output values of F gen (.), E gen (.) is a service model corresponding to the transformed face image, wherein if the output value f of F gen (.) satisfies 64<f<80, the image is determined to be User 1, that is, the output value of E gen (.) is 1, if the output value f of F gen (.) satisfies 144<f<160, the image is determined to be User 2, that is, the output value of E gen (.) is 2, if the output value f of F gen (.) falls in other ranges, the image is determined to be a non-vehicle owner user, that is, the output value of E gen (.) is 0. gen

[0083] To determine whether the transformed face image can replace the real face image, the output value of E gen (.) can be compared with the output value of E raw (.). If the error between the two is less than a preset error, it is considered that the transformed face image can replace the real face image. If the error between the two is greater than or equal to the preset error, the StyleGAN2 model is used to generate a new transformed face image again, and the above steps are repeated until the error is less than a preset range.

[0084] Specifically, for User 2, if the output value of E corresponding to one or more generated transformed face images gen (.) is always unable to be 2, it is considered that the one or more transformed face images corresponding to User 2 cannot replace the real face image to achieve the target function. At this time, it is necessary to regenerate a new transformed face image, for example, changing the input noise of the StyleGAN2 model to regenerate a face image different from the previously generated face image, and re-executing the above steps.

[0085] In summary, the method provided in the embodiments of this specification transforms the original biometric information and detects the availability of the transformed biometric information under the target service, so as to finally obtain the transformed biometric information that can achieve an effect almost similar to that of the original biometric information under the target service. After the transformed biometric information is successfully obtained, the original biometric information can be deleted, and the non-sensitive transformed biometric information can be retained. At the same time, the generated transformed biometric information is only partially similar to the original biometric information, and it is almost impossible to reverse-derive the original biometric information from the transformed biometric information, thereby effectively ensuring the user's personal data security.​

[0086] Based on the same inventive concept, embodiments of this specification also provide a bio-information conversion device, such as... Figure 4 As shown, it includes:

[0087] The first processing module 401 is used to acquire the user's original biological information and, based on the original biological information and a preset biological information generation model, generate transformed biological information corresponding to the original biological information.

[0088] The business data generation module 402 is used to train a preset business model under the target business based on the transformed biological information, and to process the transformed biological information based on the trained business model to obtain the first business data corresponding to the transformed biological information.

[0089] The second processing module 403 is used to use the transformed biological information as a substitute for the original biological information if the error between the first business data and the original biological information under the second business data of the target business is less than a preset error.

[0090] In some implementations, the first processing module 401 is used for:

[0091] Determine whether the original training sample set of the preset bioinformatics generation model contains N first bioinformatics, wherein the similarity between the features corresponding to the N first bioinformatics and the features corresponding to the original bioinformatics is greater than or equal to a preset similarity, and N is a positive integer;

[0092] If the original training sample set does not contain the N first biological information, based on the original biological information, M second biological information are determined for adjusting the preset biological information generation model, wherein the similarity between the features corresponding to the M second biological information and the features corresponding to the original biological information is greater than or equal to the preset similarity, and M is a positive integer;

[0093] The preset biological information is updated based on the M second biological information, and the transformed biological information is output.

[0094] In some implementations, the business data generation module 402 is used for:

[0095] Based on the transformed biological information, a first initial feature extraction model is trained, and based on the trained first feature extraction model, feature data of the transformed biological information is extracted.

[0096] The preset business model is trained based on the feature data of the transformed biological information.

[0097] The data output range of the first feature extraction model is the same as that of the second feature extraction model, and the second feature extraction model is a model used to extract features from the original biological information.

[0098] In some embodiments, the apparatus further includes:

[0099] The feature extraction module is used to extract feature data of the original biological information based on the second feature extraction model;

[0100] The third processing module is used to process the feature data of the original biological information based on the preset business model to obtain the second business data corresponding to the original biological information.

[0101] In some embodiments, the apparatus further includes:

[0102] The fourth processing module is used to repeatedly execute the steps of generating new transformed biological information through the preset biological information generation model and training the preset business model based on the new transformed biological information to obtain new business data corresponding to the new transformed biological information when the error between the first business data and the second business data is greater than or equal to the preset error, until the error between the obtained business data and the second business data is less than the preset error.

[0103] In some implementations, the first processing module 401 is used for:

[0104] During the validity period of the user's original biometric information, the user's original biometric information is obtained.

[0105] In some implementations, the original biometric information includes the user's real image information, and the preset biometric information generation model includes a generative adversarial network model.

[0106] The specific functions of each part of the above-mentioned device have been described in detail in the bioinformatics transformation method provided in the embodiments of this specification, and will not be elaborated here.

[0107] Based on the same inventive concept, embodiments of the present invention provide an intelligent vehicle, see reference. Figure 5 As shown, it includes a memory 1004, a processor 1002, and a computer program stored in the memory 1004 and executable on the processor 1002. When the processor 1002 executes the program, it implements any one of the embodiments of the bioinformatics transformation method.

[0108] Among them, Figure 5In this document, a bus architecture (represented by bus 1000) is used. Bus 1000 may include any number of interconnected buses and bridges, linking various circuits including one or more processors represented by processor 1002 and memory represented by memory 1004. Bus 1000 may also link various other circuits such as peripheral devices, voltage regulators, and power management circuits, which are well known in the art and therefore will not be described further herein. Bus interface 1005 provides an interface between bus 1000 and receiver 1001 and transmitter 1003. Receiver 1001 and transmitter 1003 may be the same element, i.e., a transceiver, providing a unit for communicating with various other devices over a transmission medium. Processor 1002 is responsible for managing bus 1000 and general processing, while memory 1004 may be used to store data used by processor 1002 during operation.

[0109] The functions described herein can be implemented in hardware, software executed by a processor, firmware, or any combination thereof. If implemented in software executed by a processor, the functions can be stored as one or more instructions or codes on or transmitted via a computer-readable medium. Other examples and embodiments are within the scope and spirit of this invention and the appended claims. For example, due to the nature of software, the functions described above can be implemented using software executed by a processor, hardware, firmware, hardwired, or any combination thereof. Furthermore, the functional units can be integrated into a single processing unit, or each unit can exist physically separately, or two or more units can be integrated into a single unit.

[0110] In the several embodiments provided in this application, it should be understood that the disclosed technical content can be implemented in other ways. The device embodiments described above are merely illustrative; for example, the division of units can be a logical functional division, and in actual implementation, there may be other division methods. For instance, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the displayed or discussed mutual coupling, direct coupling, or communication connection may be through some interfaces; the indirect coupling or communication connection between units or modules may be electrical or other forms.

[0111] The units described as separate components may or may not be physically separate. Similarly, the components of the control device may or may not be physical units; they may be located in one place or distributed across multiple units. Some or all of the units can be selected to achieve the purpose of this embodiment, depending on actual needs.

[0112] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, read-only memory (ROM), random access memory (RAM), portable hard drives, magnetic disks, or optical disks.

[0113] It should be noted that the above embodiments are illustrative of the invention and not restrictive, and that those skilled in the art can devise alternative embodiments without departing from the scope of the appended claims. In the claims, any reference signs placed between parentheses should not be construed as limiting the claims. The word "comprising" does not exclude the presence of elements or steps not listed in the claims. The word "a" or "an" preceding an element does not exclude the presence of a plurality of such elements. The invention can be implemented by means of hardware comprising several different elements and by means of a suitably programmed computer. In the unit claims enumerating several means, several of these means may be embodied by the same item of hardware. The use of the words first, second, and third, etc., does not indicate any order. These words can be interpreted as names.

[0114] Other embodiments of the invention will readily occur to those skilled in the art upon consideration of the specification and practice of the invention disclosed herein. This application is intended to cover any variations, uses, or adaptations of the invention that follow the general principles of the invention and include common knowledge or customary techniques in the art not disclosed herein. The specification and examples are to be considered exemplary only, and the true scope and spirit of the invention are indicated by the following claims.

[0115] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A bioinformatics transformation method, characterized in that, include: Obtain the user's original biological information, and based on the original biological information and a preset biological information generation model, generate transformed biological information corresponding to the original biological information; Based on the transformed biological information, a preset business model under the target business is trained, and the transformed biological information is processed based on the trained business model to obtain the first business data corresponding to the transformed biological information. If the error between the first business data and the original biological information in the second business data under the target business is less than a preset error, then the transformed biological information will be used as the substitute information for the original biological information. The step of generating transformed biological information corresponding to the original biological information based on the original biological information and a preset biological information generation model includes: Determine whether the original training sample set of the preset bioinformatics generation model contains N first bioinformatics, wherein the similarity between the features corresponding to the N first bioinformatics and the features corresponding to the original bioinformatics is greater than or equal to a preset similarity, and N is a positive integer; If the original training sample set does not contain the N first biological information, based on the original biological information, M second biological information are determined for adjusting the preset biological information generation model, wherein the similarity between the features corresponding to the M second biological information and the features corresponding to the original biological information is greater than or equal to the preset similarity, and M is a positive integer; The preset biological information generation model is updated based on the M second biological information, and the transformed biological information is output.

2. The method as described in claim 1, characterized in that, The step of training a preset business model for the target business based on the transformed biological information includes: Based on the transformed biological information, a first initial feature extraction model is trained, and based on the trained first feature extraction model, feature data of the transformed biological information is extracted. The preset business model is trained based on the feature data of the transformed biological information. The data output range of the first feature extraction model is the same as that of the second feature extraction model, and the second feature extraction model is a model used to extract features from the original biological information.

3. The method as described in claim 2, characterized in that, If the error between the first business data and the original biological information in the second business data under the target business is less than a preset error, before using the transformed biological information as a substitute for the original biological information, the method further includes: Based on the second feature extraction model, feature data of the original biological information are extracted; Based on the preset business model, the feature data of the original biological information are processed to obtain the second business data corresponding to the original biological information.

4. The method as described in claim 1, characterized in that, After obtaining the first business data corresponding to the transformed biological information, the method further includes: If the error between the first business data and the second business data is greater than or equal to the preset error, the steps of generating new transformed biological information through the preset biological information generation model and training the preset business model based on the new transformed biological information to obtain new business data corresponding to the new transformed biological information are repeated until the error between the obtained business data and the second business data is less than the preset error.

5. The method as described in claim 1, characterized in that, The acquisition of the user's original biometric information includes: During the validity period of the user's original biometric information, the user's original biometric information is obtained.

6. The method as described in claim 1, characterized in that, The original biometric information includes the user's real image information, and the preset biometric information generation model includes a generative adversarial network model.

7. A bio-information conversion device, characterized in that, include: The first processing module is used to acquire the user's original biological information and, based on the original biological information and a preset biological information generation model, generate transformed biological information corresponding to the original biological information. The business data generation module is used to train a preset business model under the target business based on the transformed biological information, and to process the transformed biological information based on the trained business model to obtain the first business data corresponding to the transformed biological information. The second processing module is used to use the transformed biological information as a substitute for the original biological information if the error between the first business data and the original biological information under the second business data of the target business is less than a preset error. The first processing module is configured to determine whether the original training sample set of the preset biological information generation model contains N first biological information items, wherein the similarity between the features corresponding to the N first biological information items and the features corresponding to the original biological information is greater than or equal to a preset similarity, and N is a positive integer; if the original training sample set does not contain the N first biological information items, based on the original biological information, determine M second biological information items for adjusting the preset biological information generation model, wherein the similarity between the features corresponding to the M second biological information items and the features corresponding to the original biological information is greater than or equal to the preset similarity, and M is a positive integer; update the preset biological information generation model based on the M second biological information items, and output the transformed biological information.

8. An intelligent vehicle, characterized in that, It includes a memory and one or more programs, wherein the one or more programs are stored in the memory and configured to be executed by one or more processors, containing operation instructions for performing the methods described in any one of claims 1-6.

9. A computer-readable storage medium, characterized in that, It stores a computer program that, when executed by a processor, implements the steps of the method according to any one of claims 1-6.

Images