Automatic driving method, device, processing equipment and storage medium
By encrypting some parameters of the autonomous driving neural network and generating a decryption function based on the target vehicle information, the application security problem of the neural network on unauthorized vehicles is solved, and the secure use of the neural network is realized.
Patent Information
- Application Number
- CN202110778287.0
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2021-07-09
- Publication Date
- 2024-11-26
- Estimated Expiration
- 2041-07-09
AI Technical Summary
The training process of existing autonomous driving neural networks is complex and takes a lot of time and resources, and can cause losses if the training results are applied to unauthorized vehicles.
The encryption function is used to encrypt some network parameters of the neural network, generate encryption functions based on the setting information of the target vehicle, and send the encrypted network parameters to the processing equipment of the target vehicle. After decrypting the decryption function, use the neural network for automatic driving to ensure that Only vehicles designed and trained for this neural network can be decrypted and used.
It ensures the security of the neural network, prevents unauthorized vehicles from being used, and ensures that the neural network can only be used on the target vehicles designed and trained, thereby improving the security of the application.
Smart Images

Figure CN113515812B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of autonomous driving technology, and in particular to an autonomous driving method, apparatus, processing device and storage medium. Background Art
[0002] In recent years, artificial intelligence has been deeply developed and applied in the field of autonomous driving. For example, trained neural networks can be used to obtain vehicle control information based on the information currently collected and generated by the vehicle, thereby achieving autonomous driving.
[0003] However, since the design and training process of neural networks used for autonomous driving are complex and the training process consumes a lot of time and resources, if the training results are applied to unauthorized vehicles, it will cause losses to multiple parties. Summary of the invention
[0004] In view of this, the present disclosure proposes an autonomous driving solution to solve the above technical problems.
[0005] In a first aspect, an autonomous driving method is proposed, which is applied to a server, and the method includes: obtaining a trained neural network, wherein the neural network is used to generate control information of a vehicle based on environmental perception information and vehicle status information; encrypting at least part of the network parameters of the neural network using an encryption function to obtain encrypted network parameters, wherein the encryption function is generated based on first vehicle information; sending the encrypted network parameters and indication information to a processing device of a target vehicle, so that the processing device obtains second vehicle information based on the indication information, and generates a decryption function based on the second vehicle information; when the second vehicle information is consistent with the first vehicle information, decrypting the encrypted network parameters using the decryption function to obtain decrypted network parameters, so as to perform autonomous driving using the neural network.
[0006] In combination with any embodiment provided in the present disclosure, the network parameters of the neural network include parameter information of the target vehicle.
[0007] In combination with any embodiment provided in the present disclosure, the method of using an encryption function to encrypt at least part of the network parameters of the neural network to obtain encrypted network parameters includes: using the encryption function to encrypt one layer of parameters of the network parameters to obtain encrypted network parameters.
[0008] In a second aspect, an autonomous driving method is proposed, which is applied to a processing device of a target vehicle, the method comprising: obtaining encrypted network parameters and indication information of a trained neural network, wherein the neural network is used to generate control information of the vehicle based on environmental perception information and vehicle status information, and the encrypted network parameters are obtained by encrypting at least part of the network parameters of the neural network using an encryption function, and the encryption function is generated based on the first vehicle information; obtaining second vehicle information based on the indication information, and generating a decryption function based on the second vehicle information; decrypting the encrypted network parameters using the decryption function, and obtaining the decrypted network parameters when the second vehicle information is consistent with the first vehicle information; and performing autonomous driving using the neural network.
[0009] In combination with any embodiment provided in the present disclosure, the network parameters of the neural network include parameter information of the target vehicle.
[0010] In combination with any implementation manner provided in the present disclosure, the encrypted network parameter is obtained by encrypting one layer of parameters of the network parameter.
[0011] In a third aspect, an automatic driving device is proposed, which is applied to a server, and the device includes: an acquisition unit, used to acquire a trained neural network, wherein the neural network is used to generate vehicle control information based on environmental perception information and vehicle status information; an encryption unit, used to encrypt at least part of the network parameters of the neural network using an encryption function to obtain encrypted network parameters, wherein the encryption function is generated based on first vehicle information; a sending unit, used to send the encrypted network parameters and indication information to a processing device of a target vehicle, so that the processing device obtains second vehicle information based on the indication information, and generates a decryption function based on the second vehicle information; when the second vehicle information is consistent with the first vehicle information, the encrypted network parameters are decrypted by the decryption function to obtain decrypted network parameters, so as to perform automatic driving using the neural network.
[0012] In combination with any embodiment provided in the present disclosure, the network parameters of the neural network include parameter information of the target vehicle.
[0013] In combination with any implementation manner provided in the present disclosure, the encryption unit is specifically used to: use the encryption function to encrypt one layer of parameters of the network parameters to obtain encrypted network parameters.
[0014] In a fourth aspect, an automatic driving device is proposed, which is applied to a processing device of a target vehicle, and the device includes: a receiving unit, used to obtain encrypted network parameters and indication information of a trained neural network, wherein the neural network is used to generate vehicle control information based on environmental perception information and vehicle status information, and the encrypted network parameters are obtained by encrypting at least part of the network parameters of the neural network using an encryption function, and the encryption function is generated based on the first vehicle information; an acquisition unit, used to obtain second vehicle information based on the indication information, and generate a decryption function based on the second vehicle information; a decryption unit, used to decrypt the encrypted network parameters using the decryption function, and obtain the decrypted network parameters when the second vehicle information is consistent with the first vehicle information; and a driving unit, used to perform automatic driving using the neural network.
[0015] In combination with any embodiment provided in the present disclosure, the network parameters of the neural network include parameter information of the target vehicle.
[0016] In combination with any implementation manner provided in the present disclosure, the encrypted network parameter is obtained by encrypting one layer of parameters of the network parameter.
[0017] In a fifth aspect, a processing device is proposed, which includes a memory and a processor, wherein the memory is used to store computer instructions that can be executed on the processor, and the processor is used to implement any embodiment provided by the present disclosure when executing the computer instructions.
[0018] In a sixth aspect, a computer-readable storage medium is proposed, on which a computer program is stored, and when the program is executed by a processor, any embodiment provided by the present disclosure is implemented.
[0019] In a seventh aspect, a computer program product is proposed, comprising a computer program, which, when executed by a processor, implements any embodiment provided in the present disclosure.
[0020] The automatic driving scheme provided by the present disclosure uses an encryption function to encrypt at least part of the network parameters of the trained neural network to obtain encrypted network parameters, wherein the encryption function is generated according to the setting information of the target vehicle; the encrypted network parameters are sent to the processing device of the target vehicle so that the processing device obtains the second vehicle information according to the indication information, and generates a decryption function according to the second vehicle information; when the second vehicle information is consistent with the first vehicle information, the encrypted network parameters are decrypted by the decryption function to obtain the decrypted network parameters, so as to use the neural network for automatic driving. By encrypting the network parameters of the neural network according to the first vehicle information, only the target vehicle for which the neural network is designed and trained can decrypt the encrypted network parameters, so that the neural network can be used for automatic driving, thereby ensuring the security of the neural network application.
[0021] 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
[0022] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the specification and, together with the description, serve to explain the principles of the specification.
[0023] Figure 1 A flowchart showing an autonomous driving method proposed by at least one embodiment of the present disclosure;
[0024] Figure 2 A flowchart showing another autonomous driving method proposed by at least one embodiment of the present disclosure;
[0025] Figure 3 A schematic diagram showing the structure of an automatic driving device proposed by at least one embodiment of the present disclosure is shown;
[0026] Figure 4 A schematic diagram showing the structure of another automatic driving device proposed by at least one embodiment of the present disclosure;
[0027] Figure 5 A structural diagram of a processing device proposed in at least one embodiment of the present disclosure is shown. DETAILED DESCRIPTION
[0028] The exemplary embodiments will be described in detail herein, examples of which are shown in the accompanying drawings. When the following description refers to the drawings, the same numbers in different drawings represent the same or similar elements unless otherwise indicated. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the present disclosure. Instead, they are merely examples of embodiments consistent with the appended claims.
[0029] The terms used in the present invention are only for the purpose of describing specific embodiments and are not intended to limit the present invention. The singular forms "a", "the" and "the" used in the present invention and the appended claims are also intended to include plural forms unless the context clearly indicates other meanings. It should also be understood that the term "and / or" used herein refers to and includes any or all possible combinations of one or more associated listed items.
[0030] It should be understood that although the terms first, second, etc. may be used in the present invention to describe various structures or information, these structures should not be limited to these terms. These terms are only used to distinguish structures or information of the same type from each other.
[0031] Figure 1 A flowchart of an autonomous driving method proposed by at least one embodiment of the present disclosure is shown. The method is applied to a server, which may be a server of a developer or supplier of a neural network, or a server for an autonomous driving vehicle. Figure 1 As shown, the automatic driving method may include steps 101 to 103.
[0032] In step 101, a trained neural network is obtained, wherein the neural network is used to process and generate vehicle control information according to environmental perception information and vehicle state information, that is, a neural network used for autonomous driving.
[0033] In the embodiments of the present disclosure, the environmental perception information is usually detected by sensors installed on the vehicle, including distance data, speed data, status data, etc. of various target objects in the vehicle's surrounding environment, and the sensors are, for example, cameras, laser radars, millimeter-wave radars, etc. The vehicle status information is, for example, information such as the vehicle's position, speed, and driving direction.
[0034] The neural network can be trained according to the environment perception information samples and the vehicle state information samples. The training execution subject can be the server or other devices, and the present disclosure does not limit this.
[0035] In step 102, at least part of the network parameters of the neural network are encrypted using an encryption function to obtain encrypted network parameters, wherein the encryption function is generated according to setting information of the target vehicle.
[0036] The network parameters include connection weights of neurons in each layer, bias values, etc. All network parameters can be encrypted using an encryption function, or part of the network parameters can be encrypted.
[0037] In one example, the encryption function may be used to encrypt one layer of network parameters, such as encrypting one layer of connection weights, or encrypting one layer of connection weights and bias values simultaneously, to obtain encrypted network parameters.
[0038] By encrypting one layer of connection weights, the system burden for decryption during the application process is reduced, while also ensuring the security of the neural network.
[0039] Furthermore, the encrypted network parameters may be stored as a weight file in a set format, and the set format may be the same as the storage format of the original unencrypted network parameters.
[0040] The encryption function may be selected from an encryption function library consisting of a series of encryption functions, and the encryption function library may include encryption functions of the following multiple algorithms: AES, DES, RSA, 3DES, and the like.
[0041] In the disclosed embodiment, the encryption function is generated according to the setting information of the target vehicle. The target vehicle may be a vehicle of a certain brand, or a certain model, or a certain series of vehicles, or a vehicle authorized to use the neural network, or a vehicle with a certain common feature; the setting information may be specific information unique to the target vehicle. In other words, the encryption function is designed for the target vehicle. For example, if the neural network is designed for a vehicle using an XX model engine, then the setting information of the target vehicle is the model of the engine. The encryption function is generated by the model of the engine, and at least part of the network parameters are encrypted to obtain encrypted network parameters. In step 103, the encrypted network parameters are sent to the processing device of the target vehicle so that the processing device obtains the setting information from the target vehicle to obtain a decryption function, and the encrypted network parameters are decrypted by the decryption function to obtain the decrypted network parameters, so as to use the neural network to process the environmental information and driving information to obtain the control information of the vehicle.
[0042] When the neural network is applied, the neural network model is installed in the processing device of the target vehicle. The processing device can obtain the setting information from the target vehicle and generate a decryption function according to the setting information to decrypt the encrypted network parameters to obtain the decrypted network parameters.
[0043] Taking the setting information as the model of the engine used by the vehicle as an example, during the application process, the processing device obtains the model of the engine used by its own vehicle and generates a decryption function based on the model of the engine. If the signal of the engine used by the vehicle is consistent with the engine model used by the server to generate the encryption function, the decryption function generated by the processing device can successfully decrypt the encrypted network parameters; conversely, if the model of the engine used by the own vehicle is inconsistent with the engine model used for encryption, the encrypted network parameters cannot be decrypted, that is, the neural network cannot be used for autonomous driving, thereby preventing the neural network from being stolen by unauthorized vehicles and ensuring the security of the neural network.
[0044] In an embodiment of the present disclosure, an encryption function is used to encrypt at least part of the network parameters of a trained neural network to obtain encrypted network parameters, wherein the encryption function is generated according to setting information of a target vehicle; the encrypted network parameters are sent to a processing device of the target vehicle so that the processing device obtains the setting information from the target vehicle to obtain a decryption function, and the encrypted network parameters are decrypted by the decryption function to obtain decrypted network parameters, so that the neural network is used to process environmental information and driving information to obtain vehicle control information, and the parameters of the neural network are encrypted according to the setting information of the target vehicle, so that only the target vehicle specifically designed and trained for the neural network can decrypt the encrypted network parameters, so that the neural network can be used for autonomous driving, thereby ensuring the security of the neural network application.
[0045] In some implementations, the network parameters of the neural network include parameter information of the target vehicle. That is to say, the neural network is designed for the target vehicle. Since the neural network parameters include parameter information of the target vehicle, when applied to the target vehicle, the neural network can generate accurate control information based on environmental perception information and vehicle status information. For example, the network parameters of the neural network include the internal and external parameters of the camera of a specific model of vehicle. For other different models of vehicles, due to different camera installation positions and different models, the internal and external parameters of the camera may be different from those in the network parameters. Therefore, the environmental perception information collected by the cameras of other different models of vehicles is input into the neural network, and the control information obtained may be biased. When the neural network is applied to the specific model of vehicle, since the internal and external parameters of the camera are consistent with the internal and external parameters of the camera included in the network parameters of the neural network, accurate vehicle control information can be obtained based on the environmental perception information collected by the camera of the vehicle for automatic driving.
[0046] In some implementation schemes, the server may send instruction information while sending the encrypted network parameters to the target vehicle, wherein the instruction information is used to instruct the processing device how to obtain the setting information from the target vehicle.
[0047] The indication information may indicate a specific storage address. For example, the specific storage address is used to store the engine model of the vehicle. By sending the specific storage address to the target vehicle, the target vehicle can obtain the engine model from the characteristic storage address. When the obtained engine model is consistent with the engine model used for encryption, the decryption function generated according to the obtained engine model can successfully decrypt the encrypted network parameters; conversely, when the obtained engine model is inconsistent with the engine model used for encryption, the generated decryption function cannot decrypt the encrypted network parameters, and the vehicle cannot use the neural network for autonomous driving.
[0048] Figure 2 A flowchart of another autonomous driving method proposed by at least one embodiment of the present disclosure is shown. The method is applied to a server that can be a server of a developer or supplier of a neural network, or a server for an autonomous driving vehicle. Figure 1 As shown, the automatic driving method may include steps 101 to 103.
[0049] In step 101, a trained neural network is obtained, wherein the neural network is used to process and generate vehicle control information according to environmental perception information and vehicle state information, that is, a neural network used for autonomous driving.
[0050] In the embodiments of the present disclosure, the environmental perception information is usually detected by sensors installed on the vehicle, including distance data, speed data, status data, etc. of various target objects in the vehicle's surrounding environment, and the sensors are, for example, cameras, laser radars, millimeter-wave radars, etc. The vehicle status information is, for example, information such as the vehicle's position, speed, and driving direction.
[0051] The neural network can be trained according to the environment perception information samples and the vehicle state information samples. The training execution subject can be the server or other devices, and the present disclosure does not limit this.
[0052] In step 102, at least part of the network parameters of the neural network are encrypted using an encryption function to obtain encrypted network parameters, wherein the encryption function is generated according to setting information of the target vehicle.
[0053] The network parameters include connection weights of neurons in each layer, bias values, etc. All network parameters can be encrypted using an encryption function, or part of the network parameters can be encrypted.
[0054] In one example, the encryption function can be used to encrypt network parameters.
[0055] One layer of parameters is encrypted, for example, one layer of connection weight is encrypted, or one layer of connection weight and bias value are encrypted at the same time to obtain encrypted network parameters.
[0056] By encrypting one layer of connection weights, the system burden for decryption during the application process is reduced, while also ensuring the security of the neural network.
[0057] Furthermore, the encrypted network parameters may be stored as a weight file in a set format, and the set format may be the same as the storage format of the original unencrypted network parameters.
[0058] The encryption function may be selected from an encryption function library consisting of a series of encryption functions, and the encryption function library may include encryption functions of the following multiple algorithms: AES, DES, RSA, 3DES, and the like.
[0059] In the disclosed embodiment, the encryption function is generated according to the setting information of the target vehicle. The target vehicle may be a vehicle of a certain brand, or a certain model, or a certain series of vehicles, or a vehicle authorized to use the neural network, or a vehicle with a certain common feature; the setting information may be specific information unique to the target vehicle. In other words, the encryption function is designed for the target vehicle. For example, if the neural network is designed for a vehicle using an XX model engine, then the setting information of the target vehicle is the model of the engine. The encryption function is generated by the model of the engine, and at least part of the network parameters are encrypted to obtain encrypted network parameters. In step 103, the encrypted network parameters are sent to the processing device of the target vehicle so that the processing device obtains the setting information from the target vehicle to obtain a decryption function, and the encrypted network parameters are decrypted by the decryption function to obtain the decrypted network parameters, so as to use the neural network to process the environmental information and driving information to obtain the control information of the vehicle.
[0060] When the neural network is applied, the neural network model is installed in the processing device of the target vehicle. The processing device can obtain the setting information from the target vehicle and generate a decryption function according to the setting information to decrypt the encrypted network parameters to obtain the decrypted network parameters.
[0061] Taking the setting information as the model of the engine used by the vehicle as an example, during the application process, the processing device obtains the model of the engine used by its own vehicle and generates a decryption function based on the model of the engine. If the signal of the engine used by the vehicle is consistent with the engine model used by the server to generate the encryption function, the decryption function generated by the processing device can successfully decrypt the encrypted network parameters; conversely, if the model of the engine used by the own vehicle is inconsistent with the engine model used for encryption, the encrypted network parameters cannot be decrypted, that is, the neural network cannot be used for autonomous driving, thereby preventing the neural network from being stolen by unauthorized vehicles and ensuring the security of the neural network.
[0062] In an embodiment of the present disclosure, an encryption function is used to encrypt at least part of the network parameters of a trained neural network to obtain encrypted network parameters, wherein the encryption function is generated according to setting information of a target vehicle; the encrypted network parameters are sent to a processing device of the target vehicle so that the processing device obtains the setting information from the target vehicle to obtain a decryption function, and the encrypted network parameters are decrypted by the decryption function to obtain decrypted network parameters, so that the neural network is used to process environmental information and driving information to obtain vehicle control information, and the parameters of the neural network are encrypted according to the setting information of the target vehicle, so that only the target vehicle specifically designed and trained for the neural network can decrypt the encrypted network parameters, so that the neural network can be used for autonomous driving, thereby ensuring the security of the neural network application.
[0063] In some implementations, the network parameters of the neural network include parameter information of the target vehicle. That is to say, the neural network is designed for the target vehicle. Since the neural network parameters include parameter information of the target vehicle, when applied to the target vehicle, the neural network can generate accurate control information based on environmental perception information and vehicle status information. For example, the network parameters of the neural network include the internal and external parameters of the camera of a specific model of vehicle. For other different models of vehicles, due to different camera installation positions and different models, the internal and external parameters of the camera may be different from those in the network parameters. Therefore, the environmental perception information collected by the cameras of other different models of vehicles is input into the neural network, and the control information obtained may be biased. When the neural network is applied to the specific model of vehicle, since the internal and external parameters of the camera are consistent with the internal and external parameters of the camera included in the network parameters of the neural network, accurate vehicle control information can be obtained based on the environmental perception information collected by the camera of the vehicle for automatic driving.
[0064] In some implementation schemes, the server may send instruction information while sending the encrypted network parameters to the target vehicle, wherein the instruction information is used to instruct the processing device how to obtain the setting information from the target vehicle.
[0065] The indication information may indicate a specific storage address. For example, the specific storage address is used to store the engine model of the vehicle. By sending the specific storage address to the target vehicle, the target vehicle can obtain the engine model from the characteristic storage address. When the obtained engine model is consistent with the engine model used for encryption, the decryption function generated according to the obtained engine model can successfully decrypt the encrypted network parameters; conversely, when the obtained engine model is inconsistent with the engine model used for encryption, the generated decryption function cannot decrypt the encrypted network parameters, and the vehicle cannot use the neural network for autonomous driving.
[0066] Figure 3 A schematic structural diagram of an automatic driving device proposed by at least one embodiment of the present disclosure is shown, and the device is applied to a server, and the device may include: an acquisition unit 301, used to acquire a trained neural network, wherein the neural network is used to generate vehicle control information based on environmental perception information and vehicle status information; an encryption unit 302, used to encrypt at least part of the network parameters of the neural network using an encryption function to obtain encrypted network parameters, wherein the encryption function is generated based on first vehicle information; a sending unit 303, used to send the encrypted network parameters and indication information to a processing device of a target vehicle, so that the processing device obtains second vehicle information based on the indication information, and generates a decryption function based on the second vehicle information; when the second vehicle information is consistent with the first vehicle information, the encrypted network parameters are decrypted by the decryption function to obtain decrypted network parameters, so as to perform automatic driving using the neural network.
[0067] In some implementations, the network parameters of the neural network include parameter information of the target vehicle.
[0068] In some implementations, the encryption unit is specifically used to: use the encryption function to encrypt one layer of parameters of the network parameters to obtain encrypted network parameters.
[0069] Figure 4A schematic structural diagram of another automatic driving device proposed by at least one embodiment of the present disclosure is shown, and the device is applied to a processing device of a target vehicle, and the device may include: a receiving unit 401, used to obtain encrypted network parameters and indication information of a trained neural network, wherein the neural network is used to generate vehicle control information based on environmental perception information and vehicle status information, and the encrypted network parameters are obtained by encrypting at least part of the network parameters of the neural network using an encryption function, and the encryption function is generated based on the first vehicle information; an acquisition unit 402, used to obtain second vehicle information based on the indication information, and generate a decryption function based on the second vehicle information; a decryption unit 403, used to decrypt the encrypted network parameters using the decryption function, and obtain the decrypted network parameters when the second vehicle information is consistent with the first vehicle information; a driving unit 404, used to perform automatic driving using the neural network.
[0070] In some implementations, the network parameters of the neural network include parameter information of the target vehicle.
[0071] In some implementations, the encrypted network parameter is obtained by encrypting one layer of parameters of the network parameter.
[0072] At least one embodiment of the present disclosure further provides a processing device, such as Figure 5 As shown, the device includes a memory and a processor, the memory is used to store computer instructions that can be executed on the processor, and the processor is used to implement the automatic driving method described in any embodiment of the present disclosure when executing the computer instructions.
[0073] At least one embodiment of the present disclosure further provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the autonomous driving method described in any embodiment of the present disclosure.
[0074] At least one embodiment of the present disclosure further provides a computer program product, including a computer program, which, when executed by a processor, implements the autonomous driving method described in any embodiment of the present disclosure.
[0075] It should be understood by those skilled in the art that one or more embodiments of the present specification may be provided as a method, system or computer program product. Therefore, one or more embodiments of the present specification may take the form of a complete hardware embodiment, a complete software embodiment or an embodiment combining software and hardware. Moreover, one or more embodiments of the present specification may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.
[0076] Each embodiment in this specification is described in a progressive manner, and the same or similar parts between the embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the data processing device embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiment.
[0077] The above is a description of a specific embodiment of the specification. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recorded in the claims can be performed in an order different from that in the embodiments and still achieve the desired results. In addition, the processes depicted in the drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.
[0078] The embodiments of the subject matter and functional operations described in this specification may be implemented in the following: digital electronic circuits, tangibly embodied computer software or firmware, computer hardware including the structures disclosed in this specification and their structural equivalents, or a combination of one or more of them. The embodiments of the subject matter described in this specification may be implemented as one or more computer programs, i.e., one or more modules in computer program instructions encoded on a tangible non-temporary program carrier to be executed by a data processing device or to control the operation of the data processing device. Alternatively or additionally, the program instructions may be encoded on an artificially generated propagation signal, such as a machine-generated electrical, optical or electromagnetic signal, which is generated to encode information and transmit it to a suitable receiver device for execution by a data processing device. The computer storage medium may be a machine-readable storage device, a machine-readable storage substrate, a random or serial access memory device, or a combination of one or more of them.
[0079] The processes and logic flows described in this specification can be performed by one or more programmable computers executing one or more computer programs to perform corresponding functions by operating on input data and generating output. The processes and logic flows can also be performed by special purpose logic circuits, such as FPGAs (field programmable gate arrays) or ASICs (application-specific integrated circuits), and the apparatus can also be implemented as special purpose logic circuits.
[0080] Computers suitable for executing computer programs include, for example, general and / or special microprocessors, or any other type of central processing unit. Typically, the central processing unit will receive instructions and data from a read-only memory and / or a random access memory. The basic components of a computer include a central processing unit for implementing or executing instructions and one or more memory devices for storing instructions and data. Typically, the computer will also include one or more large-capacity storage devices for storing data, such as magnetic disks, magneto-optical disks, or optical disks, or the computer will be operably coupled to this large-capacity storage device to receive data from it or to transmit data to it, or both. However, the computer does not necessarily have such a device. In addition, the computer can be embedded in another device, such as a mobile phone, a personal digital assistant (PDA), a mobile audio or video player, a game console, a global positioning system (GPS) receiver, or a portable storage device such as a universal serial bus (USB) flash drive, to name a few.
[0081] Computer readable media suitable for storing computer program instructions and data include all forms of non-volatile memory, media and memory devices, including, for example, semiconductor memory devices (e.g., EPROM, EEPROM and flash memory devices), magnetic disks (e.g., internal hard disks or removable disks), magneto-optical disks, and CD ROM and DVD-ROM disks. The processor and memory may be supplemented by, or incorporated in, special purpose logic circuitry.
[0082] Although this specification includes many specific implementation details, these should not be interpreted as limiting the scope of any invention or the scope of protection claimed, but are mainly used to describe the features of the specific embodiments of specific inventions. Certain features described in multiple embodiments in this specification may also be implemented in combination in a single embodiment. On the other hand, the various features described in a single embodiment may also be implemented separately in multiple embodiments or in any suitable sub-combination. In addition, although features may work in certain combinations as described above and even initially claim protection, one or more features from the claimed combination may be removed from the combination in some cases, and the claimed combination may point to a sub-combination or a variation of a sub-combination.
[0083] Similarly, although operations are depicted in a particular order in the accompanying drawings, this should not be understood as requiring that these operations be performed in the particular order shown or performed sequentially, or requiring that all illustrated operations be performed to achieve the desired results. In some cases, multitasking and parallel processing may be advantageous. In addition, the separation of various system modules and components in the above-described embodiments should not be understood as requiring such separation in all embodiments, and it should be understood that the described program components and systems can generally be integrated together in a single software product, or packaged into multiple software products.
[0084] Thus, specific embodiments of the subject matter have been described. Other embodiments are within the scope of the appended claims. In some cases, the actions recited in the claims can be performed in a different order and still achieve the desired results. In addition, the processes depicted in the drawings do not necessarily require the particular order or sequential order shown to achieve the desired results. In some implementations, multitasking and parallel processing may be advantageous.
[0085] The above description is merely a preferred embodiment of one or more embodiments of the present specification and is not intended to limit one or more embodiments of the present specification. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of one or more embodiments of the present specification shall be included in the scope of protection of one or more embodiments of the present specification.
Claims
1. An automatic driving method, characterized in that: Applied to a server, the method comprises: Acquire a trained neural network, wherein the neural network is used to generate vehicle control information based on environmental perception information and vehicle state information; Encrypting at least part of the network parameters of the neural network using an encryption function to obtain encrypted network parameters, wherein the encryption function is generated according to first vehicle information of a set vehicle; The encrypted network parameters and indication information are sent to a processing device of a target vehicle, so that the processing device obtains second vehicle information according to the indication information and generates a decryption function according to the second vehicle information; when the second vehicle information is consistent with the first vehicle information, the encrypted network parameters are decrypted by the decryption function to obtain decrypted network parameters, so as to use the neural network for autonomous driving, wherein the neural network is installed in the processing device of the target vehicle.
2. The method according to claim 1, characterized in that The network parameters of the neural network include parameter information of the set vehicle.
3. The method according to claim 1 or 2, characterized in that: The step of encrypting at least part of the network parameters of the neural network using an encryption function to obtain encrypted network parameters includes: The encryption function is used to encrypt one layer of parameters of the network parameters to obtain encrypted network parameters.
4. An automatic driving method, characterized in that: A processing device applied to a target vehicle, the method comprising: Obtaining encrypted network parameters and indication information of a trained neural network, wherein the neural network is used to generate vehicle control information based on environmental perception information and vehicle state information, and the encrypted network parameters are obtained by encrypting at least part of the network parameters of the neural network using an encryption function, and the encryption function is generated based on the first vehicle information; Acquire second vehicle information according to the indication information, and generate a decryption function according to the second vehicle information; decrypting the encrypted network parameters using the decryption function, and obtaining the decrypted network parameters when the second vehicle information is consistent with the first vehicle information; The neural network is used for autonomous driving.
5. The method according to claim 4, characterized in that The network parameters of the neural network include parameter information for setting the vehicle.
6. The method according to claim 4 or 5, characterized in that: The encrypted network parameter is obtained by encrypting one layer of parameters of the network parameter.
7. An automatic driving device, characterized in that: Applied to a server, the device comprises: an acquisition unit, configured to acquire a trained neural network, wherein the neural network is configured to generate vehicle control information based on environmental perception information and vehicle state information; an encryption unit, configured to encrypt at least part of the network parameters of the neural network using an encryption function to obtain encrypted network parameters, wherein the encryption function is generated according to the first vehicle information; A sending unit is used to send the encrypted network parameters and indication information to a processing device of a target vehicle, so that the processing device obtains second vehicle information according to the indication information and generates a decryption function according to the second vehicle information; when the second vehicle information is consistent with the first vehicle information, the encrypted network parameters are decrypted by the decryption function to obtain decrypted network parameters, so as to use the neural network for automatic driving, wherein the neural network is installed in the processing device of the target vehicle.
8. An automatic driving device, characterized in that: A processing device applied to a target vehicle, the device comprising: a receiving unit, configured to obtain encrypted network parameters and indication information of a trained neural network, wherein the neural network is configured to generate vehicle control information based on environmental perception information and vehicle state information, and the encrypted network parameters are obtained by encrypting at least part of the network parameters of the neural network using an encryption function, and the encryption function is generated based on the first vehicle information; an acquisition unit, configured to acquire second vehicle information according to the indication information, and generate a decryption function according to the second vehicle information; a decryption unit, configured to decrypt the encrypted network parameters using the decryption function, and obtain the decrypted network parameters when the second vehicle information is consistent with the first vehicle information; A driving unit is used for performing automatic driving using the neural network.
9. A processing device, characterized in that: The device includes a memory and a processor, wherein the memory is used to store computer instructions executable on the processor, and the processor is used to implement any one of the methods described in claims 1 to 3 when executing the computer instructions, or to implement any one of the methods described in claims 4 to 6.
10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the method described in any one of claims 1 to 3 is implemented, or the method described in any one of claims 4 to 6 is implemented.
