Data processing method and related apparatus

Abstract

The embodiments of the present application provide a data processing method. When a simulated traffic scenario is generated, the requirements for the simulated scenario to be generated can be acquired. The requirements, for example, may comprise first information and second information, wherein the first information comprises a road structure of a driving scenario, and the second information comprises the category of a driving behavior of a target vehicle in the driving scenario. Feature representations of the first information and the second information are used to generate a simulated scenario of the driving scenario by means of a generative model, wherein the simulated scenario may comprise movement information of a plurality of objects, including the target vehicle, in the driving scenario. In the embodiments of the present application, on the basis of the description (for example, comprising a road structure of a driving scenario and the category of a driving behavior of a host vehicle) from a user for a simulated scenario to be generated, the simulated scenario is generated by means of a generative model, so that a simulated scenario that satisfies the description of the simulated scenario to be generated can be obtained, and a plurality of types of driving scenarios can be generalized by means of the generative model, thereby obtaining more diversified simulated scenarios.

Description

A data processing method and related apparatus

[0001] This application claims priority to Chinese Patent Application No. 202410564430.X, filed on April 30, 2024, entitled “A Data Processing Method and Related Apparatus”, the entire contents of which are incorporated herein by reference. Technical Field

[0002] This application relates to the field of artificial intelligence, and more particularly to a data processing method and apparatus thereof. Background Technology

[0003] Autonomous driving simulation is of great significance for the testing and training of control systems. Autonomous vehicles often encounter dangerous, highly interactive scenarios caused by dynamic obstacles during operation. These scenarios need to be reproduced in simulation for autonomous driving algorithm optimization and testing, or for training decision-making networks. However, such scenarios are relatively few, and replayed scenarios lack interactive capabilities. Therefore, it is necessary to generalize and augment these scenarios, fit and learn their distribution, generate similar simulation scenarios, and enable key dynamic obstacles to have real-time interactive capabilities.

[0004] In existing technologies, when generating simulation scenarios, road test data is collected and recorded using autonomous driving or other vehicles. Key scenario information contained in the road test records is extracted, and behavioral models are trained using machine learning and other methods based on the historical trajectories of obstacles collected during road tests. These models are then stored in a behavioral model library. In other words, behavioral models of different obstacles can be pre-generated. For the simulation scenario to be generated, the most matching behavioral model is searched from the pre-generated behavioral models. However, this approach relies on the pre-generated behavioral models of obstacles and cannot construct new scenarios based on the original scenario (i.e., scenarios not covered by the behavioral models). This limits the diversity of the generated simulation scenarios. Summary of the Invention

[0005] In a first aspect, this application provides a data processing method that, when generating a simulated traffic scene, can obtain the requirements of the simulated scene to be generated, such as first information and second information. The first information includes the road structure of the driving scene, and the second information includes the category of driving behavior of the target vehicle in the driving scene. Using the feature representation of the first information and the second information, a simulation scene of the driving scene is generated through a generative model. The simulation scene can include the movement information of multiple objects, including the target vehicle, in the driving scene.

[0006] The object can be the target vehicle (i.e., the main vehicle) or other dynamic obstacles.

[0007] In existing technologies, behavioral models of different obstacles are pre-generated. For the simulation scenario to be generated, the most matching behavioral model is searched from the pre-generated behavioral models. This method relies on the pre-generated behavioral models of obstacles and cannot construct new scenarios based on the original scenario (i.e., scenarios not covered by the behavioral models), which limits the diversity of the generated simulation scenarios. In the embodiments of this application, the simulation scenario is generated by generating a model based on the user's description of the generated simulation scenario (e.g., including the road structure of the driving scenario and the driving behavior type of the main vehicle). This can yield simulation scenarios that meet the description of the generated simulation scenario. The generation model can be generalized to various types of driving scenarios, resulting in more diverse simulation scenarios.

[0008] The road structure in the driving scenario can be described as road condition information, which may include the topological structure (e.g., location or shape) of traffic objects (e.g., lane lines, obstacles, etc.). Formally, the road structure can be represented by a map. The target vehicle in the simulation scenario obtained through the generative model interacts with other objects among the plurality of objects to perform the aforementioned category of driving behavior.

[0009] This allows you to specify the types of driving behaviors the target vehicle needs to perform in the generated simulation scenario. These driving behaviors can be interactions with other obstacles (such as other vehicles). For example, the types of driving behaviors can be: obstacle avoidance, lane keeping, lane changing, cutting in, and merging into traffic.

[0010] In this embodiment, the parameters of the model (e.g., encoder or generator model) can be adjusted based on the effect of the downstream training task, thereby obtaining a simulation scenario that can achieve better results in the downstream training task. Specifically, in one possible implementation, the training effect obtained by training the machine learning model according to the simulation scenario can also be obtained, and the encoder or generator model can be updated with the goal of optimizing the training effect.

[0011] During the training process of a model, the model parameters need to be updated based on the set training objectives. For example, the training objective can be to improve the model's processing accuracy for data in a specific scenario or to improve the processing accuracy for data in a generalized scenario. The training effect can be the degree to which the updated model, obtained by training the model based on the training samples, meets the training objective. For example, the degree of improvement in the processing accuracy for data in a specific scenario or the degree of improvement in the processing accuracy for data in a generalized scenario.

[0012] In one possible implementation, the second information specifically refers to obtaining a feature representation through an encoder, including obtaining the feature representation through an encoder based on a prompt containing the second information.

[0013] In one possible implementation, the encoder includes a first encoder and a second encoder. Based on first information, a feature representation of the first information is obtained through the first encoder; based on second information, a feature representation of the second information is obtained through the second encoder. This allows for the acquisition of the training effect obtained from training the machine learning model based on the simulation scenario, and the second encoder is updated with the goal of optimizing the training effect. Specifically, when encoding the second information, a feature representation can be obtained through the encoder (e.g., the second encoder) based on a prompt containing the second information. Updating the second encoder is equivalent to using a soft prompt scheme to fine-tune the guiding prompt, thereby obtaining a more effective simulation scenario for downstream tasks.

[0014] In one possible implementation, the second encoder is a transformer-based neural network model.

[0015] In one possible implementation, the second information also includes the target vehicle's driving style.

[0016] In one possible implementation, the generative model can be updated based on the relationship between movement information and kinematic constraints. For example, a suitable reward model can be used to constrain the scene generation effect and quality, ensuring that the scene meets actual driving constraints. Kinematic constraints could include Jerk constraints, collision constraints, etc. When updating the generative model, a reward tuning algorithm can be used to guide the generated model to satisfy predetermined constraints (i.e., kinematic constraints) by updating the model with rewards.

[0017] In one possible implementation, the generative model is a diffusion model.

[0018] Secondly, this application provides a data processing apparatus, the apparatus comprising:

[0019] The acquisition module is used to acquire first information and second information. The first information includes the road structure of the driving scenario, and the second information includes the category of the target vehicle's driving behavior in the driving scenario.

[0020] The processing module is used to obtain feature representations through an encoder based on the first information and the second information; and to obtain a simulation scene of the driving scenario through a generation model based on the feature representations. The simulation scene includes the movement information of multiple objects, including the target vehicle, in the driving scene, and the target vehicle in the simulation scene performs driving behaviors of a certain category.

[0021] In one possible implementation, the driving behavior is categorized as follows:

[0022] Vehicle obstacle avoidance, lane keeping, lane changing, lane cutting, and merging into traffic.

[0023] In one possible implementation, the processing module is also used for:

[0024] Obtain the training results of training a machine learning model based on a simulation scenario;

[0025] The goal is to optimize training results by updating the encoder or generating the model.

[0026] In one possible implementation, the second information specifically refers to the feature representation obtained through the encoder, including:

[0027] Based on the prompt containing the second information, the encoder obtains the feature representation.

[0028] In one possible implementation, the encoder includes a first encoder and a second encoder;

[0029] The processing module is specifically used for:

[0030] Based on the first information, the feature representation of the first information is obtained through the first encoder;

[0031] Based on the second information, the feature representation of the second information is obtained through the second encoder;

[0032] Update the second encoder.

[0033] In one possible implementation, the second encoder is a transformer-based neural network model.

[0034] In one possible implementation, the second information also includes the target vehicle's driving style.

[0035] In one possible implementation, the processing module is also used for:

[0036] The generative model is updated based on the relationship between the movement information and the kinematic constraints.

[0037] In one possible implementation, the generative model is a diffusion model.

[0038] Thirdly, embodiments of this application provide a data processing apparatus, which may include a memory, a processor, and a bus system, wherein the memory is used to store a program, and the processor is used to execute the program in the memory to perform the methods described in the first aspect above and any of its optional methods.

[0039] Fourthly, embodiments of this application provide a computer-readable storage medium storing a computer program that, when run on a computer, causes the computer to perform the methods described in the first aspect and any of its optional methods.

[0040] Fifthly, embodiments of this application provide a computer program that, when run on a computer, causes the computer to perform the first aspect and any of its optional methods described above.

[0041] Sixthly, this application provides a chip system including a processor for supporting an execution device or training device in implementing the functions involved in the foregoing aspects, such as transmitting or processing data involved in the foregoing methods; or, information. In one possible design, the chip system further includes a memory for storing program instructions and data necessary for the execution device or training device. This chip system may be composed of chips or may include chips and other discrete devices. Attached Figure Description

[0042] Figure 1A is a schematic diagram of a structural framework for artificial intelligence.

[0043] Figures 1B and 1C are schematic diagrams of the application system framework of the present invention;

[0044] Figure 1D is a schematic diagram of an optional hardware structure for the terminal;

[0045] Figure 2 is a schematic diagram of a server structure;

[0046] Figure 3 is a schematic diagram of a system architecture according to this application;

[0047] Figure 4 illustrates the process of a cloud service.

[0048] Figure 5 is a flowchart illustrating a data processing method provided in an embodiment of this application;

[0049] Figure 6 is a flowchart illustrating a data processing method provided in an embodiment of this application;

[0050] Figures 7 and 8 illustrate the correction of the predicted solution;

[0051] Figure 9 is a schematic diagram of the application system framework of the present invention;

[0052] Figure 10 is a schematic diagram of a data processing device provided in an embodiment of this application;

[0053] Figure 11 is a schematic diagram of a terminal device provided in an embodiment of this application;

[0054] Figure 12 is a schematic diagram of a server structure provided in an embodiment of this application;

[0055] Figure 13 is a schematic diagram of a chip structure provided in an embodiment of this application. Detailed Implementation

[0056] The embodiments of the present invention will now be described with reference to the accompanying drawings. The terminology used in the embodiments section is for illustrative purposes only and is not intended to limit the scope of the invention.

[0057] The embodiments of this application will now be described with reference to the accompanying drawings. Those skilled in the art will recognize that, with technological advancements and the emergence of new scenarios, the technical solutions provided in the embodiments of this application are equally applicable to similar technical problems.

[0058] The terms "first," "second," etc., used in the specification, claims, and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such terms are interchangeable where appropriate; this is merely a way of distinguishing objects with the same attributes in the embodiments of this application. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion, so that a process, method, system, product, or apparatus that comprises a series of elements is not necessarily limited to those elements, but may include other elements not explicitly listed or inherent to those processes, methods, products, or apparatuses.

[0059] The terms “substantially,” “about,” and similar terms used herein are used as approximations rather than as terms of degree, and are intended to take into account the inherent biases of measurements or calculations known to those skilled in the art. Furthermore, the use of “may” in describing embodiments of the invention refers to “one or more possible embodiments.” The terms “use,” “using,” and “used” used herein are to be considered synonymous with the terms “utilize,” “utilizing,” and “utilized,” respectively. Additionally, the term “exemplary” is intended to refer to an instance or illustration.

[0060] First, the overall workflow of an artificial intelligence system is described, as shown in Figure 1A. Figure 1A is a structural diagram of the main framework of artificial intelligence. The framework is then elaborated from two dimensions: the "intelligent information chain" (horizontal axis) and the "IT value chain" (vertical axis). The "intelligent information chain" reflects a series of processes from data acquisition to processing. For example, it could be the general process of intelligent information perception, intelligent information representation and formation, intelligent reasoning, intelligent decision-making, and intelligent execution and output. In this process, data undergoes a condensation process of "data—information—knowledge—wisdom." The "IT value chain" reflects the value that artificial intelligence brings to the information technology industry, from the underlying infrastructure of human intelligence and information (provided and processed by technology) to the industrial ecosystem of the system.

[0061] (1) Infrastructure

[0062] Infrastructure provides computing power to support artificial intelligence systems, enabling communication with the external world and providing support through a basic platform. This communication occurs through sensors; computing power is provided by intelligent chips (hardware acceleration chips such as CPUs, NPUs, GPUs, ASICs, and FPGAs); and the basic platform includes distributed computing frameworks and related platform guarantees and support, which may include cloud storage and computing, interconnected networks, etc. For example, sensors communicate with the outside world to acquire data, and this data is provided to intelligent chips in the distributed computing system provided by the basic platform for computation.

[0063] (2) Data

[0064] The data at the next layer of infrastructure is used to represent the data sources in the field of artificial intelligence. The data involves graphics, images, voice, text, and IoT data from traditional devices, including business data from existing systems and sensor data such as force, displacement, liquid level, temperature, and humidity.

[0065] (3) Data processing

[0066] Data processing typically includes methods such as data training, machine learning, deep learning, search, reasoning, and decision-making.

[0067] Among them, machine learning and deep learning can perform intelligent information modeling, extraction, preprocessing, and training on data, including symbolization and formalization.

[0068] Reasoning refers to the process in which, in a computer or intelligent system, the machine thinks and solves problems by simulating human intelligent reasoning, based on reasoning control strategies and using formalized information. Typical functions include search and matching.

[0069] Decision-making refers to the process of making decisions based on intelligent information after reasoning, and it typically provides functions such as classification, sorting, and prediction.

[0070] (4) General ability

[0071] After the data processing mentioned above, the results of the data processing can be used to form some general capabilities, such as algorithms or a general system, for example, translation, text analysis, computer vision processing, speech recognition, image recognition, etc.

[0072] (5) Smart Products and Industry Applications

[0073] Intelligent products and industry applications refer to products and applications of artificial intelligence systems in various fields. They are the encapsulation of overall artificial intelligence solutions, productizing intelligent information decision-making and realizing practical applications. Their application areas mainly include: intelligent terminals, intelligent transportation, intelligent healthcare, autonomous driving, smart cities, etc.

[0074] First, we will introduce the application scenarios of this application. This application can be used in, but is not limited to, applications with generative artificial intelligence (AIGC) functionality or cloud services provided by cloud-side servers. These will be described in detail below:

[0075] I. Data generation (e.g., training sample generation) applications

[0076] The product form of this application embodiment can be a data generation (e.g., training sample generation) application. Data generation (e.g., training sample generation) applications can run on terminal devices or cloud-based servers.

[0077] In one possible implementation, a data generation (e.g., training sample generation) application can perform a data generation (e.g., training sample generation) task based on input data, wherein the data generation (e.g., training sample generation) application can perform the data generation (e.g., training sample generation) task in response to input data and obtain the generated result.

[0078] For example, the training samples generated from the data in this application embodiment can be training samples for an autonomous driving model. These training samples can be simulation results of driving scenarios (also referred to as simulation scenarios in this application embodiment). The simulation scenario can include motion information of multiple vehicles in the driving environment.

[0079] In one possible implementation, a user can open a data generation (e.g., training sample generation) application installed on a terminal device and input data. The data generation (e.g., training sample generation) application can generate data (e.g., training sample generation) from the input data using the methods provided in the embodiments of this application, and present the generated results to the user (the presentation method may include, but is not limited to, displaying, saving, uploading to the cloud, etc.).

[0080] In one possible implementation, a user can open a data generation (e.g., training sample generation) application installed on a terminal device and input data. The data generation (e.g., training sample generation) application can send the input data to a cloud-based server. The cloud-based server uses the method provided in this application embodiment to generate data (e.g., training sample generation) from the input data and sends the generation result back to the terminal device. The terminal device can then present the generation result to the user (the presentation method may include, but is not limited to, displaying, saving, uploading to the cloud, etc.).

[0081] The following sections will introduce the data generation (e.g., training sample generation) application class in the embodiments of this application from the perspectives of functional architecture and product architecture that implements the functions.

[0082] Referring to Figure 1B, Figure 1B is a schematic diagram of the functional architecture of a data generation (e.g., training sample generation) application in an embodiment of this application:

[0083] In one possible implementation, as shown in FIG1B, a data generation (e.g., training sample generation) application 102 may receive input parameters 101 (e.g., containing input data) and generate a result 103. The data generation (e.g., training sample generation) application 102 may execute on at least one computer system (for example) and includes computer code that, when executed by one or more computers, causes the computers to execute a natural language model trained by the methods provided in the embodiments of this application.

[0084] Referring to Figure 1C, Figure 1C is a schematic diagram of the entity architecture of a data generation (e.g., training sample generation) application in an embodiment of this application:

[0085] Referring to Figure 1C, which illustrates a system architecture diagram, the system may include a terminal 100 and a server 200. The server 200 may include one or more servers (Figure 1C illustrates this using one server as an example), and the server 200 may provide data generation (e.g., training sample generation) services to one or more terminals.

[0086] The terminal 100 may have a data generation (e.g., training sample generation) application installed, or a webpage related to the data generation (e.g., training sample generation) function opened. The application and webpage can provide an interface. The terminal 100 can receive relevant parameters input by the user on the data generation (e.g., training sample generation) function interface and send the parameters to the server 200. The server 200 can obtain the processing result based on the received parameters and return the processing result to the terminal 100.

[0087] It should be understood that in some optional implementations, the terminal 100 can also complete the action of obtaining the processing result based on the received parameters on its own, without the need for the server to cooperate. This application embodiment is not limited to this.

[0088] The product form of terminal 100 in Figure 1C will be described next;

[0089] The terminal 100 in this application embodiment can be a mobile phone, tablet computer, wearable device, vehicle device, augmented reality (AR) / virtual reality (VR) device, laptop computer, ultra-mobile personal computer (UMPC), netbook, personal digital assistant (PDA), etc., and this application embodiment does not impose any restrictions on it.

[0090] Figure 1D shows a schematic diagram of an optional hardware structure for terminal 100.

[0091] Referring to Figure 1D, terminal 100 may include components such as a radio frequency unit 110, a memory 120, an input unit 130, a display unit 140, a camera 150 (optional), an audio circuit 160 (optional), a speaker 161 (optional), a microphone 162 (optional), a processor 170, an external interface 180, and a power supply 190. Those skilled in the art will understand that Figure 1D is merely an example of a terminal or multi-functional device and does not constitute a limitation on the terminal or multi-functional device; it may include more or fewer components than illustrated, or combine certain components, or use different components.

[0092] The input unit 130 can be used to receive input numerical or character information, and to generate key signal inputs related to user settings and function control of the portable multi-functional device. Specifically, the input unit 130 may include a touchscreen 131 (optional) and / or other input devices 132. The touchscreen 131 can collect touch operations performed by the user on or near it (such as operations performed by the user using fingers, knuckles, styluses, or any suitable object on or near the touchscreen), and drive the corresponding connection devices according to a pre-set program. The touchscreen can detect the user's touch actions, convert the touch actions into touch signals and send them to the processor 170, and can receive and execute commands sent by the processor 170; the touch signal includes at least touch point coordinate information. The touchscreen 131 can provide an input interface and an output interface between the terminal 100 and the user. In addition, various types of touchscreens, such as resistive, capacitive, infrared, and surface acoustic wave, can be used to implement the touchscreen. Besides the touchscreen 131, the input unit 130 may also include other input devices. Specifically, other input devices 132 may include, but are not limited to, one or more of the following: physical keyboard, function keys (such as volume control buttons 132, power buttons 133, etc.), trackball, mouse, joystick, etc.

[0093] Among them, the input device 132 can receive input data, etc.

[0094] The display unit 140 can be used to display information input by the user or information provided to the user, various menus of the terminal 100, interactive interfaces, file display, and / or playback of any multimedia file. In this embodiment, the display unit 140 can be used to display the interface of a data generation (e.g., training sample generation) application, the generated results, etc.

[0095] The memory 120 can be used to store instructions and data. The memory 120 may primarily include an instruction storage area and a data storage area. The data storage area can store various types of data, such as multimedia files and text. The instruction storage area can store software units such as operating systems, applications, and instructions required for at least one function, or subsets or extended sets thereof. It may also include non-volatile random access memory. It provides the processor 170 with hardware, software, and data resources for managing the computing device, supporting control software and applications. It is also used for storing multimedia files, as well as storing running programs and applications.

[0096] The processor 170 is the control center of the terminal 100. It connects various parts of the terminal 100 via various interfaces and lines. By running or executing instructions stored in the memory 120 and calling data stored in the memory 120, it performs various functions and processes data of the terminal 100, thereby controlling the terminal device as a whole. Optionally, the processor 170 may include one or more processing units; preferably, the processor 170 may integrate an application processor and a modem processor, wherein the application processor mainly handles the operating system, user interface, and applications, and the modem processor mainly handles wireless communication. It is understood that the modem processor may not be integrated into the processor 170. In some embodiments, the processor and memory can be implemented on a single chip; in some embodiments, they can also be implemented separately on independent chips. The processor 170 can also be used to generate corresponding operation control signals, send them to the corresponding components of the computing processing device, read and process data in the software, especially read and process data and programs in the memory 120, so that the various functional modules therein perform corresponding functions, thereby controlling the corresponding components to act according to the instructions.

[0097] The memory 120 can be used to store software code related to the data processing method, and the processor 170 can execute the steps of the chip's data processing method, and can also schedule other units (such as the above-mentioned input unit 130 and display unit 140) to achieve the corresponding functions.

[0098] The radio frequency unit 110 (optional) can be used for receiving and transmitting signals during information transmission or calls. For example, it can receive downlink information from the base station and process it for the processor 170; additionally, it can transmit uplink data to the base station. Typically, the RF circuit includes, but is not limited to, an antenna, at least one amplifier, a transceiver, a coupler, a low-noise amplifier (LNA), a duplexer, etc. Furthermore, the radio frequency unit 110 can also communicate wirelessly with network devices and other devices. This wireless communication can use any communication standard or protocol, including but not limited to Global System for Mobile communication (GSM), General Packet Radio Service (GPRS), Code Division Multiple Access (CDMA), Wideband Code Division Multiple Access (WCDMA), Long Term Evolution (LTE), email, Short Messaging Service (SMS), etc.

[0099] In this embodiment of the application, the radio frequency unit 110 can send input data to the server 200 and receive the generation result sent by the server 200.

[0100] It should be understood that the radio frequency unit 110 is optional and can be replaced with other communication interfaces, such as a network port.

[0101] The terminal 100 also includes a power supply 190 (such as a battery) that supplies power to various components. Preferably, the power supply can be logically connected to the processor 170 through a power management system, thereby enabling functions such as charging, discharging, and power consumption management through the power management system.

[0102] Terminal 100 also includes an external interface 180, which can be a standard Micro USB interface or a multi-pin connector, which can be used to connect terminal 100 to other devices for communication or to connect a charger to charge terminal 100.

[0103] Although not shown, terminal 100 may also include a flash, a wireless fidelity (WiFi) module, a Bluetooth module, and sensors with various functions, which will not be described in detail here. Some or all of the methods described below can be applied to terminal 100 as shown in Figure 1D.

[0104] The product form of server 200 in Figure 1C is described below;

[0105] Figure 2 provides a schematic diagram of the structure of a server 200. As shown in Figure 2, the server 200 includes a bus 201, a processor 202, a communication interface 203, and a memory 204. The processor 202, the memory 204, and the communication interface 203 communicate with each other via the bus 201.

[0106] Bus 201 can be a Peripheral Component Interconnect (PCI) bus or an Extended Industry Standard Architecture (EISA) bus, etc. Buses can be categorized as address buses, data buses, control buses, etc. For ease of representation, only one thick line is used in Figure 2, but this does not indicate that there is only one bus or one type of bus.

[0107] The processor 202 can be any one or more of the following processors: central processing unit (CPU), graphics processing unit (GPU), microprocessor (MP), or digital signal processor (DSP).

[0108] Memory 204 may include volatile memory, such as random access memory (RAM). Memory 204 may also include non-volatile memory, such as read-only memory (ROM), flash memory, hard disk drive (HDD), or solid state drive (SSD).

[0109] The memory 204 can be used to store software code related to the data processing method, and the processor 202 can execute the steps of the chip's data processing method, and can also schedule other units to achieve corresponding functions.

[0110] It should be understood that the aforementioned terminal 100 and server 200 can be centralized or distributed devices. The processors (e.g., processor 170 and processor 202) in the aforementioned terminal 100 and server 200 can be hardware circuits (such as application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), general-purpose processors, digital signal processors (DSPs), microprocessors or microcontrollers, etc.) or combinations of these hardware circuits. For example, the processor can be a hardware system with instruction execution capabilities, such as a CPU or DSP, or a hardware system without instruction execution capabilities, such as an ASIC or FPGA, or a combination of the aforementioned hardware systems without instruction execution capabilities and hardware systems with instruction execution capabilities.

[0111] It should be understood that the steps related to the model inference process in the embodiments of this application involve AI-related operations. When performing AI operations, the instruction execution architecture of the terminal device and the server is not limited to the processor-memory architecture described above. The system architecture provided in the embodiments of this application will be described in detail below with reference to Figure 3.

[0112] Figure 3 is a schematic diagram of the system architecture provided in an embodiment of this application. As shown in Figure 3, the system architecture 500 includes an execution device 510, a training device 520, a database 530, a client device 540, a data storage system 550, and a data acquisition system 560.

[0113] The execution device 510 includes a calculation module 511, an I / O interface 512, a preprocessing module 513, and a preprocessing module 514. The calculation module 511 may include a target model / rule 501, while the preprocessing modules 513 and 514 are optional.

[0114] The execution device 510 can be a terminal device or server for the aforementioned application that generates running data (e.g., training sample generation).

[0115] The data acquisition device 560 is used to collect training samples. Training samples can be program files (including program code and program input data), etc. After collecting the training samples, the data acquisition device 560 stores these training samples in the database 530.

[0116] The training device 520 can maintain training samples in the database 530 to obtain the target model / rule 501 from the neural network to be trained.

[0117] It should be noted that in practical applications, the training samples maintained in database 530 may not all come from the data acquisition device 560; they may also be received from other devices. Furthermore, it should be noted that training device 520 may not necessarily train the target model / rule 501 entirely based on the training samples maintained in database 530; it may also obtain training samples from the cloud or other sources for model training. The above description should not be construed as limiting the embodiments of this application.

[0118] The target model / rule 501 trained by the training device 520 can be applied to different systems or devices, such as the execution device 510 shown in Figure 3. The execution device 510 can be a terminal, such as a mobile terminal, tablet computer, laptop computer, augmented reality (AR) / virtual reality (VR) device, vehicle terminal, etc., or it can be a server, etc.

[0119] Specifically, the training device 520 can transfer the trained model to the execution device 510.

[0120] In Figure 3, the execution device 510 is configured with an input / output (I / O) interface 512 for data interaction with external devices. Users can input data to the I / O interface 512 through the client device 540 (e.g., input data in the embodiment of this application).

[0121] Preprocessing modules 513 and 514 are used to preprocess the input data received from the I / O interface 512. It should be understood that preprocessing modules 513 and 514 may be absent, or only one preprocessing module may be used. When preprocessing modules 513 and 514 are absent, the calculation module 511 can be used directly to process the input data.

[0122] During the preprocessing of input data by the execution device 510, or during the calculation module 511 of the execution device 510 performing calculations and other related processes, the execution device 510 can call data, code, etc. in the data storage system 550 for corresponding processing, or store the data, instructions, etc. obtained from the corresponding processing into the data storage system 550.

[0123] Finally, the I / O interface 512 provides the processing results (such as the generated results) to the client device 540, thereby providing them to the user.

[0124] In the scenario shown in Figure 3, the user can manually provide input data, which can be done through the interface provided by I / O interface 512. Alternatively, the client device 540 can automatically send input data to I / O interface 512. If user authorization is required for the client device 540 to automatically send input data, the user can set the corresponding permissions in the client device 540. The user can view the output results of the execution device 510 on the client device 540, which can be presented in various forms such as display, sound, or animation. The client device 540 can also act as a data acquisition terminal, collecting the input data and output results of the input I / O interface 512 as shown in the figure, and storing them as new sample data in database 530. Alternatively, data can be collected directly from the I / O interface 512 without going through the client device 540, using the input data and output results of the input I / O interface 512 as shown in the figure, and storing them as new sample data in database 530.

[0125] It is worth noting that Figure 3 is merely a schematic diagram of a system architecture provided in an embodiment of this application. The positional relationships between the devices, components, modules, etc., shown in the figure do not constitute any limitation. For example, in Figure 3, the data storage system 550 is an external memory relative to the execution device 510. In other cases, the data storage system 550 can also be placed in the execution device 510. It should be understood that the aforementioned execution device 510 can be deployed in the client device 540.

[0126] From the inference side of the model:

[0127] In this embodiment, the computing module 511 of the execution device 510 can obtain the code stored in the data storage system 550 to implement the steps related to the model reasoning process in this embodiment.

[0128] In this embodiment of the application, the computing module 511 of the execution device 510 may include hardware circuits (such as application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), general-purpose processors, digital signal processors (DSPs), microprocessors or microcontrollers, etc.) or combinations of these hardware circuits. For example, the training device 520 may be a hardware system with instruction execution capabilities, such as a CPU or DSP, or a hardware system without instruction execution capabilities, such as an ASIC or FPGA, or a combination of the aforementioned hardware systems without instruction execution capabilities and hardware systems with instruction execution capabilities.

[0129] Specifically, the computing module 511 of the execution device 510 can be a hardware system with the function of executing instructions. The steps related to the model inference process provided in this application embodiment can be software code stored in the memory. The computing module 511 of the execution device 510 can obtain the software code from the memory and execute the obtained software code to implement the steps related to the model inference process provided in this application embodiment.

[0130] It should be understood that the computing module 511 of the execution device 510 can be a combination of a hardware system without the function of executing instructions and a hardware system with the function of executing instructions. Some steps related to the model reasoning process provided in the embodiments of this application can also be implemented by the hardware system in the computing module 511 of the execution device 510 without the function of executing instructions, which is not limited here.

[0131] From the training side of the model:

[0132] In this embodiment, the training device 520 can obtain the code stored in the memory (not shown in Figure 3, which can be integrated into the training device 520 or deployed separately from the training device 520) to implement the steps related to model training in this embodiment.

[0133] In this embodiment of the application, the training device 520 may include hardware circuits (such as application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), general-purpose processors, digital signal processors (DSPs), microprocessors or microcontrollers, etc.) or combinations of these hardware circuits. For example, the training device 520 may be a hardware system with instruction execution capabilities, such as a CPU or DSP, or a hardware system without instruction execution capabilities, such as an ASIC or FPGA, or a combination of the aforementioned hardware systems without instruction execution capabilities and hardware systems with instruction execution capabilities.

[0134] It should be understood that the training device 520 can be a combination of a hardware system without the function of executing instructions and a hardware system with the function of executing instructions. Some steps related to the training of the neutralization model provided in the embodiments of this application can also be implemented by the hardware system in the training device 520 without the function of executing instructions, which is not limited here.

[0135] II. Cloud services providing data generation (e.g., training sample generation) functionality:

[0136] In one possible implementation, the server can provide data generation (e.g., training sample generation) services to the client side through an application programming interface (API).

[0137] In this process, the terminal device can send relevant parameters (such as input data) to the server through the API provided by the cloud. The server can obtain the processing result (such as the generated result) based on the received parameters and return the processing result to the terminal.

[0138] The description of the terminal and server can be found in the above embodiments, and will not be repeated here.

[0139] Figure 4 illustrates the process of using a cloud service that provides data generation (e.g., training sample generation) functionality from a cloud platform.

[0140] 1. Activate and purchase data generation (e.g., training sample generation) services.

[0141] 2. Users can download the software development kit (SDK) corresponding to the data generation (e.g., training sample generation) service. Cloud platforms typically provide multiple development versions of the SDK for users to choose from based on their development environment needs, such as a Java version SDK, a Python version SDK, a PHP version SDK, an Android version SDK, etc.

[0142] 3. After downloading the corresponding version of the SDK to their local machine according to their needs, users can import the SDK project into their local development environment, configure and debug it in the local development environment, and develop other functions in the local development environment, thus forming an application that integrates data generation (such as training sample generation) functional capabilities.

[0143] 4. When an application needs to perform data generation (e.g., training sample generation), it can trigger an API call for this function. When the application triggers this function, it sends an API request to the running instance of the data generation service in the cloud environment. The API request carries input data, which is then processed by the running instance in the cloud environment to obtain the processing result (e.g., generated result).

[0144] 5. The cloud environment returns the processing results to the application, thus completing a data generation (e.g., training sample generation) function service call.

[0145] Since the embodiments of this application involve a large number of neural network applications, for ease of understanding, the relevant terms and concepts such as neural networks involved in the embodiments of this application will be introduced below.

[0146] (1) Neural Network

[0147] A neural network can be composed of neural units, which can be defined as a computational unit that takes xs (i.e., input data) and an intercept of 1 as input. The output of this computational unit can be:

[0148] Where s = 1, 2, ..., n, where n is a natural number greater than 1, Ws is the weight of xs, and b is the bias of the neural unit. f is the activation function of the neural unit, used to introduce nonlinear characteristics into the neural network to convert the input signal in the neural unit into an output signal. The output signal of this activation function can be used as the input of the next convolutional layer, and the activation function can be the sigmoid function. A neural network is a network formed by connecting multiple of the above-mentioned individual neural units together, that is, the output of one neural unit can be the input of another neural unit. The input of each neural unit can be connected to the local receptive field of the previous layer to extract the features of the local receptive field, which can be a region composed of several neural units.

[0149] (2) Deep Neural Networks

[0150] Deep Neural Networks (DNNs), also known as multilayer neural networks, can be understood as neural networks with many hidden layers, though there's no specific metric for "many." DNNs can be categorized into three layers based on their position: input layers, hidden layers, and output layers. Generally, the first layer is the input layer, the last layer is the output layer, and the layers in between are hidden layers. All layers are fully connected, meaning that any neuron in the i-th layer is connected to any neuron in the (i+1)-th layer. Although DNNs appear complex, the operation of each layer is actually quite simple, resembling a linear relationship as follows: in, It is the input vector. It is the output vector. α is the offset vector, W is the weight matrix (also called coefficients), and α() is the activation function. Each layer is simply an adjustment of the input vector. The output vector is obtained through such a simple operation. Because DNNs have many layers, the coefficients W and the offset vector... The number of these parameters is therefore quite large. The definitions of these parameters in a DNN are as follows: Taking the coefficient W as an example: Assuming a three-layer DNN, the linear coefficient from the 4th neuron in the second layer to the 2nd neuron in the third layer is defined as... The superscript 3 represents the layer number where coefficient W resides, while the subscript corresponds to the output third layer index 2 and the input second layer index 4. In summary, the coefficients from the k-th neuron in layer L-1 to the j-th neuron in layer L are defined as follows: It's important to note that the input layer does not have a W parameter. In deep neural networks, more hidden layers allow the network to better represent complex real-world situations. Theoretically, the more parameters a model has, the higher its complexity and "capacity," meaning it can perform more complex learning tasks. Training a deep neural network is essentially the process of learning the weight matrix, with the ultimate goal of obtaining the weight matrix of all layers in the trained deep neural network (a weight matrix formed by the vectors W from many layers).

[0151] (3) Loss Function

[0152] In training a deep neural network, to ensure the output closely approximates the desired predicted value, we compare the network's prediction with the target value. Based on the difference, we update the weight vector of each layer (usually pre-configuring parameters before the initial update). For example, if the prediction is too high, the weight vector is adjusted to predict a lower value. This adjustment continues until the deep neural network predicts the target value or a value very close to it. Therefore, we need to predefine "how to compare the difference between the predicted and target values," which is the loss function or objective function. These are important equations used to measure the difference between the predicted and target values. Taking the loss function as an example, a higher output value (loss) indicates a greater difference, and training the deep neural network becomes a process of minimizing this loss.

[0153] (4) Backpropagation algorithm

[0154] Backpropagation (BP) can be used during training to correct the parameters in the initial model, thereby reducing the model's error loss. Specifically, forward propagation of the input signal to the output generates error loss; this error loss information is then propagated back to update the parameters in the initial model, leading to convergence of the error loss. The backpropagation algorithm is an error-loss-driven backpropagation process aimed at obtaining optimal model parameters, such as the weight matrix.

[0155] (5) Diffusion model

[0156] A diffusion model is a generative model used to generate data such as images and text. The core idea of ​​a diffusion model is to diffuse noise into the data and then gradually remove the noise to recover the original data. The diffusion model consists of two stages: a forward process (noise diffusion) and a reverse process (noise removal and recovery).

[0157] Autonomous driving simulation is of great significance for the testing and training of control systems. Autonomous vehicles often encounter dangerous, highly interactive scenarios caused by dynamic obstacles during operation. These scenarios need to be reproduced in simulation for autonomous driving algorithm optimization and testing, or for training decision-making networks. However, such scenarios are relatively few, and replayed scenarios lack interactive capabilities. Therefore, it is necessary to generalize and augment these scenarios, fit and learn their distribution, generate similar simulation scenarios, and enable key dynamic obstacles to have real-time interactive capabilities.

[0158] In existing technologies, when generating simulation scenarios, road test data is collected and recorded using autonomous driving or other vehicles. Key scenario information contained in the road test records is extracted, and behavioral models are trained using machine learning and other methods based on the historical trajectories of obstacles collected during road tests. These models are then stored in a behavioral model library. In other words, behavioral models of different obstacles can be pre-generated. For the simulation scenario to be generated, the most matching behavioral model is searched from the pre-generated behavioral models. However, this approach relies on the pre-generated behavioral models of obstacles and cannot construct new scenarios based on the original scenario (i.e., scenarios not covered by the behavioral models). This limits the diversity of the generated simulation scenarios.

[0159] To address the aforementioned problems, embodiments of this application provide a data processing method. The model training method of this application embodiment will be described in detail below with reference to the accompanying drawings.

[0160] Referring to Figure 5, which is a flowchart of a data processing method provided in an embodiment of this application, the data processing method provided in this application may include steps 501 to 503, which will be described in detail below.

[0161] 501. Obtain first information and second information. The first information includes the road structure of the driving scenario, and the second information includes the category of the target vehicle's driving behavior in the driving scenario.

[0162] In one possible implementation, to generate a simulated traffic scene (which can be simply referred to as a simulated scene), the first piece of information can be obtained. This first piece of information is the road structure (i.e., environmental information, or road condition information) in the driving scene. This first piece of information can be, but is not limited to, map information, static obstacle information, and motion information of other traffic participants in the original scene. For example, the road structure can include information such as the position, positional relationship, and shape of multiple traffic units (e.g., lane lines, traffic lights, etc.).

[0163] In one possible implementation, the driving behavior type of the master vehicle (i.e., the target vehicle) in the generated simulated traffic scene can be specified. The generated simulated traffic scene can serve as a training sample for an autonomous driving model (i.e., a machine learning model used for autonomous driving tasks). The autonomous driving model determines the driving decisions of the master vehicle based on current environmental information (such as road conditions, obstacles, etc.). The simulated traffic scene can include the movement information of the target vehicle (which can be called the master vehicle) in the driving environment (e.g., driving speed, driving position, etc.). When training the machine learning model, the movement information of the target vehicle can be used as the learning target to update the model.

[0164] In one possible implementation, the category of driving behavior that the target vehicle needs to perform in the generated simulation scenario can be specified. This driving behavior can be an interaction with other obstacles (such as other vehicles). For example, the category of driving behavior can be one of the following: vehicle obstacle avoidance, lane keeping, vehicle lane changing, vehicle cutting in, and merging into traffic.

[0165] In one possible implementation, the driving style of the target vehicle in the generated simulation scenario can be specified. For example, aggressive, conservative, etc.

[0166] 502. Based on the first and second information, the feature representation is obtained through the encoder.

[0167] In one possible implementation, when encoding the second information, the feature representation can be obtained by an encoder (e.g., a second encoder) based on a prompt containing the second information.

[0168] In this embodiment of the application, by introducing a Prompt, semantic-level guidance can be provided for the generated scene, and the data distribution can be selected in a targeted manner, so that the generated scene is more in line with the expected distribution.

[0169] In one possible implementation, the encoder includes a first encoder and a second encoder; based on first information, a feature representation of the first information is obtained through the first encoder; based on second information, a feature representation of the second information is obtained through the second encoder.

[0170] In one possible implementation, the second encoder is a transformer-based neural network model. By modeling environmental information using the Transformer base model, a more accurate feature representation of the scene can be obtained.

[0171] 503. Based on the feature representation, a simulation scenario of the driving scene is obtained through the generation model. The simulation scenario includes the movement information of multiple objects, including the target vehicle, in the driving scene, and the target vehicle in the simulation scenario performs driving behaviors of a certain category.

[0172] The generative model can be a diffusion model. In the architecture of the diffusion model, the noise-adding module adds noise to the input feature representation multiple times (i.e., noise adding at multiple steps). The noise-adding module of the diffusion model can predict the noise added at each step and perform denoising based on the predicted noise to obtain the generated image. The model can be a neural network with a UNet architecture. UNet is an end-to-end convolutional neural network mainly used for image segmentation. Its characteristic is a U-shaped structure, containing an encoder and a decoder. The encoder uses convolutional layers and max-pooling layers to achieve feature extraction and downsampling; the decoder performs feature fusion and recovery through upsampling and skip connections. This architecture is mainly used to predict the noise added in the pre-pass of the diffusion framework. The embodiments of this application utilize the fitting ability of the diffusion model to complex multimodal distributions to learn the complex distribution of human and vehicle, and with the Prompt function, it can cover a wider variety of distributions.

[0173] The object can be the target vehicle (i.e., the main vehicle) or other dynamic obstacles.

[0174] In existing technologies, behavior models of different obstacles are pre-generated, and the most matching behavior model is searched from the pre-generated behavior models for the simulation scene to be generated. This method relies on the pre-generated behavior models of obstacles and cannot construct new scenes based on the original scene (that is, scenes not covered by the behavior models), which limits the diversity of the generated simulation scenes.

[0175] In this embodiment of the application, the simulation scene is generated by generating a model based on the user's description of the generated simulation scene (e.g., the road structure of the driving scene and the driving behavior type of the main vehicle). This can result in a simulation scene that satisfies the description of the generated simulation scene. The generation model can be generalized to generate various types of driving scenes, resulting in more diverse simulation scenes.

[0176] After obtaining the simulation scene, it can be converted into the format required by downstream tasks and saved for easy access by other modules.

[0177] The obtained simulation scenarios can be used to perform downstream tasks, including but not limited to: 1. Iterative optimization and testing of autonomous driving algorithms; 2. Verification of daily monitoring of autonomous driving algorithms; 3. Training of deep learning-based autonomous driving decision-making algorithms. Different downstream tasks have different optimization objectives. For example, in iterative optimization and testing of autonomous driving algorithms, the optimization objective is to maximize the model's processing accuracy for various driving behavior types. Specifically, the tasks used in simulation scenarios typically include monitoring set construction and algorithm verification. Another category is training of the control decision network, used to improve the performance of the decision network in weak scenarios and increase the pass rate of the decision network.

[0178] In this embodiment, the encoder parameters can be adjusted based on the effect of the downstream training task to obtain a simulation scenario that can achieve better results for the downstream training task. Specifically, in one possible implementation, the training effect obtained by training the machine learning model based on the simulation scenario can also be obtained, and the encoder can be updated with the goal of optimizing the training effect.

[0179] For example, to improve the passability of deep learning-based autonomous driving decision-making schemes in weak scenarios, the passability can be fed back as a downstream task indicator to the simulation scenario generation algorithm.

[0180] In one possible implementation, the encoder includes a first encoder and a second encoder. Based on first information, a feature representation of the first information is obtained through the first encoder; based on second information, a feature representation of the second information is obtained through the second encoder. This allows for the acquisition of the training effect obtained from training the machine learning model based on the simulation scenario, and the second encoder is updated with the goal of optimizing the training effect. Specifically, when encoding the second information, a feature representation can be obtained through the encoder (e.g., the second encoder) based on a prompt containing the second information. Updating the second encoder is equivalent to using a soft prompt scheme to fine-tune the guiding prompt, thereby obtaining a more effective simulation scenario for downstream tasks.

[0181] In one possible implementation, the generative model can be updated based on the relationship between movement information and kinematic constraints. For example, a suitable reward model can be used to constrain the scene generation effect and quality, ensuring that the scene meets actual driving constraints. Kinematic constraints could include Jerk constraints, collision constraints, etc. When updating the generative model, a reward tuning algorithm can be used to guide the generated model to satisfy predetermined constraints (i.e., kinematic constraints) by updating the model with rewards.

[0182] The application framework of this application embodiment will be described below with reference to the accompanying drawings:

[0183] Referring to Figure 6, which is a schematic diagram of a functional module architecture according to an embodiment of this application, it includes:

[0184] The basic scene constraint module S1 can perform the action of "updating and generating model S4 according to the relationship between the movement information and the kinematic constraints" in the above embodiment.

[0185] Scene environment information S2, this module corresponds to the acquisition of first information described in the above embodiments.

[0186] The scene guidance module S3 can perform the actions of obtaining the second information and updating the second encoder as described in the above embodiments.

[0187] Simulation scenario S5 can organize the basic scene set of the generated model into the data format required by downstream task S6 for use by downstream tasks, such as post-processing and filtering.

[0188] Evaluation module S7 can perform the actions described in the above embodiments for the effect on the downstream training task to adjust the encoder parameters.

[0189] Referring to Figure 7, which illustrates the application architecture when the driving behavior type is obstacle avoidance, the prompt can include the user's desired scenario type (e.g., the category of the target vehicle's driving behavior). For example, during training, manually labeled data is used to classify scenario types into obstacle avoidance, lane keeping, merging, and cutting in. During inference, the desired scenario type is manually input; in this embodiment, it is obstacle avoidance. Environmental information, such as obstacle history trajectory information and road topology information, is modeled using the transformer base model to obtain a complete scenario description. The entire generative model algorithm flowchart is shown in Figure 8, where the basic scenario constraints i include: trajectory smoothing, collision avoidance, and satisfying kinematic acceleration constraints. This embodiment uses reward tuning to add constraint guidance to the diffusion model, maximizing the reward value of the algorithm-generated trajectory during diffusion model training. In closed-loop downstream tasks, soft-prompt technology can be used to encode the downstream task results into suitable prompt information. The fine-tuning process only requires adjusting the soft prompt parameter; optimization of the generative model parameters is not required. In this embodiment, the downstream task feedback result is the passage rate in the weak scenario, that is, the pass rate in the obstacle avoidance scenario. The flowchart of the closed-loop downstream task can be shown in Figure 9.

[0190] Referring to Figure 10, which is a schematic diagram of the structure of a data processing apparatus provided in an embodiment of this application, as shown in Figure 10, the data processing apparatus 1000 provided in this embodiment of the application includes:

[0191] The acquisition module 1001 is used to acquire first information and second information. The first information includes the road structure of the driving scenario, and the second information includes the category of the target vehicle's driving behavior in the driving scenario.

[0192] For a detailed description of the acquisition module 1001, please refer to the description of the embodiment corresponding to Figure 5 in the above embodiments. The similarities will not be repeated here.

[0193] The processing module 1002 is used to obtain feature representations through an encoder based on the first information and the second information; and to obtain a simulation scene of the driving scenario through a generation model based on the feature representations. The simulation scene includes the movement information of multiple objects, including the target vehicle, in the driving scene, and the target vehicle in the simulation scene performs a category of driving behavior.

[0194] For a detailed description of the processing module 1002, please refer to the description of the embodiment corresponding to Figure 5 in the above embodiments. The similarities will not be repeated here.

[0195] In one possible implementation, the driving behavior is categorized as follows:

[0196] Vehicle obstacle avoidance, lane keeping, lane changing, lane cutting, and merging into traffic.

[0197] In one possible implementation, the processing module 1002 is further configured to:

[0198] Obtain the training results of training a machine learning model based on a simulation scenario;

[0199] The encoder is updated with the goal of optimizing training results.

[0200] In one possible implementation, the second information specifically refers to the feature representation obtained through the encoder, including:

[0201] Based on the prompt containing the second information, the encoder obtains the feature representation.

[0202] In one possible implementation, the encoder includes a first encoder and a second encoder;

[0203] Processing module 1002 is specifically used for:

[0204] Based on the first information, the feature representation of the first information is obtained through the first encoder;

[0205] Based on the second information, the feature representation of the second information is obtained through the second encoder;

[0206] Update the second encoder.

[0207] In one possible implementation, the second encoder is a transformer-based neural network model.

[0208] In one possible implementation, the second information also includes the target vehicle's driving style.

[0209] In one possible implementation, the processing module 1002 is further configured to:

[0210] The generative model is updated based on the relationship between the movement information and the kinematic constraints.

[0211] In one possible implementation, the generative model is a diffusion model.

[0212] The following describes an execution device provided in an embodiment of this application. Please refer to Figure 11, which is a structural schematic diagram of an execution device provided in an embodiment of this application. The execution device 1100 can specifically be a virtual reality (VR) device, a mobile phone, a tablet, a laptop, a smart wearable device, a monitoring data processing device, or a server, etc., and is not limited here. Specifically, the execution device 1100 includes: a receiver 1101, a transmitter 1102, a processor 1103, and a memory 1104 (the number of processors 1103 in the execution device 1100 can be one or more; Figure 11 shows one processor as an example). The processor 1103 may include an application processor 11031 and a communication processor 11032. In some embodiments of this application, the receiver 1101, transmitter 1102, processor 1103, and memory 1104 can be connected via a bus or other means.

[0213] Memory 1104 may include read-only memory and random access memory, and provides instructions and data to processor 1103. A portion of memory 1104 may also include non-volatile random access memory (NVRAM). Memory 1104 stores processor and operation instructions, executable modules, or data structures, or subsets thereof, or extended sets thereof, wherein the operation instructions may include various operation instructions for implementing various operations.

[0214] Processor 1103 controls the operation of the execution device. In specific applications, the various components of the execution device are coupled together through a bus system, which may include not only the data bus, but also power buses, control buses, and status signal buses. However, for clarity, all buses are referred to as the bus system in the diagram.

[0215] The methods disclosed in the embodiments of this application can be applied to or implemented by the processor 1103. The processor 1103 can be an integrated circuit chip with signal processing capabilities. During implementation, each step of the above method can be completed by the integrated logic circuitry in the hardware of the processor 1103 or by instructions in software form. The processor 1103 can be a general-purpose processor, a digital signal processor (DSP), a microprocessor, or a microcontroller, and may further include an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components. The processor 1103 can implement or execute the methods, steps, and logic block diagrams disclosed in the embodiments of this application. The general-purpose processor can be a microprocessor or any conventional processor. The steps of the methods disclosed in the embodiments of this application can be directly embodied in the execution of a hardware decoding processor, or executed by a combination of hardware and software modules in the decoding processor. The software module can reside in a mature storage medium in the field, such as random access memory, flash memory, read-only memory, programmable read-only memory, electrically erasable programmable memory, or registers. This storage medium is located in memory 1104. Processor 1103 reads the information from memory 1104 and, in conjunction with its hardware, completes the steps involved in the model inference process described above.

[0216] Receiver 1101 can be used to receive input digital or character information, and to generate signal inputs related to the settings and function control of the execution device. Transmitter 1102 can be used to output digital or character information through the first interface; transmitter 1102 can also be used to send instructions to the disk group through the first interface to modify the data in the disk group; transmitter 1102 may also include a display device such as a display screen.

[0217] This application embodiment also provides a server device. Please refer to Figure 12. Figure 12 is a schematic diagram of a server structure provided in this application embodiment. Specifically, the server 1200 is implemented by one or more servers. The server 1200 can vary significantly due to different configurations or performance. It may include one or more central processing units (CPUs) 1212 (e.g., one or more processors) and memory 1232, and one or more storage media 1230 (e.g., one or more mass storage devices) for storing application programs 1242 or data 1244. The memory 1232 and storage media 1230 can be temporary or persistent storage. The program stored in the storage media 1230 may include one or more modules (not shown in the figure), and each module may include a series of instruction operations on the server. Furthermore, the CPU 1212 may be configured to communicate with the storage media 1230 and execute the series of instruction operations in the storage media 1230 on the server 1200.

[0218] Server 1200 may also include one or more power supplies 1226, one or more wired or wireless network interfaces 1250, one or more input / output interfaces 1258; or, one or more operating systems 1241, such as Windows Server™, Mac OS X™, Unix™, Linux™, FreeBSD™, etc.

[0219] In this embodiment, the central processing unit 1212 is used to execute the data processing method described in the above embodiment.

[0220] This application also provides a computer program product that, when run on a computer, causes the computer to perform steps as performed by the aforementioned execution device, or causes the computer to perform steps as performed by the aforementioned training device.

[0221] This application also provides a computer-readable storage medium storing a program for signal processing, which, when run on a computer, causes the computer to perform steps as performed by the aforementioned execution device, or causes the computer to perform steps as performed by the aforementioned training device.

[0222] The execution device, training device, or terminal device provided in this application embodiment can specifically be a chip. The chip includes a processing unit and a communication unit. The processing unit can be, for example, a processor, and the communication unit can be, for example, an input / output interface, pins, or circuits. The processing unit can execute computer execution instructions stored in the storage unit to cause the chip within the execution device to execute the data processing method described in the above embodiments, or to cause the chip within the training device to execute the data processing method described in the above embodiments. Optionally, the storage unit can be a storage unit within the chip, such as a register or cache. Alternatively, the storage unit can be a storage unit located outside the chip within the wireless access device, such as a read-only memory (ROM) or other types of static storage devices capable of storing static information and instructions, such as random access memory (RAM).

[0223] Specifically, please refer to Figure 13, which is a schematic diagram of a chip structure provided in an embodiment of this application. The chip can be represented as a neural network processor (NPU) 1300. The NPU 1300 is mounted as a coprocessor on the host CPU, and tasks are assigned by the host CPU. The core part of the NPU is the arithmetic circuit 1303, which is controlled by the controller 1304 to extract matrix data from the memory and perform multiplication operations.

[0224] In some implementations, the arithmetic circuit 1303 internally includes multiple processing engines (PEs). In some implementations, the arithmetic circuit 1303 is a two-dimensional pulsating array. The arithmetic circuit 1303 can also be a one-dimensional pulsating array or other electronic circuits capable of performing mathematical operations such as multiplication and addition. In some implementations, the arithmetic circuit 1303 is a general-purpose matrix processor.

[0225] For example, suppose we have an input matrix A, a weight matrix B, and an output matrix C. The arithmetic circuit retrieves the corresponding data of matrix B from the weight memory 1302 and caches it in each PE of the arithmetic circuit. The arithmetic circuit retrieves the data of matrix A from the input memory 1301 and performs matrix operations with matrix B. The partial result or the final result of the obtained matrix is ​​stored in the accumulator 1308.

[0226] Unified memory 1306 is used to store input and output data. Weight data is directly transferred to weight memory 1302 via Direct Memory Access Controller (DMAC) 1305. Input data is also transferred to unified memory 1306 via DMAC.

[0227] BIU stands for Bus Interface Unit, which is used for interaction between the AXI bus and the DMAC and the Instruction Fetch Buffer (IFB) 1309.

[0228] The Bus Interface Unit (BIU) 1310 is used by the instruction fetch memory 1309 to fetch instructions from external memory, and also by the memory access controller 1305 to fetch the original data of the input matrix A or the weight matrix B from external memory.

[0229] The DMAC is mainly used to move input data from external memory DDR to unified memory 1306, or to weight data to weight memory 1302, or to input data to input memory 1301.

[0230] The vector computation unit 1307 includes multiple processing units that further process the output of the computation circuit 1303 when needed, such as vector multiplication, vector addition, exponential operations, logarithmic operations, size comparisons, etc. It is mainly used for computation in non-convolutional / fully connected layers of neural networks, such as Batch Normalization, pixel-level summation, and upsampling of feature planes.

[0231] In some implementations, the vector computation unit 1307 can store the processed output vector in the unified memory 1306. For example, the vector computation unit 1307 can apply a linear function, or a nonlinear function, to the output of the computation circuit 1303, such as performing linear interpolation on the feature planes extracted by the convolutional layer, or, for example, accumulating a vector of values ​​to generate activation values. In some implementations, the vector computation unit 1307 generates normalized values, pixel-level summed values, or both. In some implementations, the processed output vector can be used as an activation input to the computation circuit 1303, for example, for use in subsequent layers of the neural network.

[0232] The instruction fetch buffer 1309 connected to the controller 1304 is used to store the instructions used by the controller 1304;

[0233] Unified memory 1306, input memory 1301, weighted memory 1302, and instruction fetch memory 1309 are all on-chip memories. External memory is proprietary to this NPU hardware architecture.

[0234] The processor mentioned above can be a general-purpose central processing unit, a microprocessor, an ASIC, or one or more integrated circuits used to control the execution of the above program.

[0235] It should also be noted that the device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. In addition, in the device embodiment drawings provided in this application, the connection relationship between modules indicates that they have a communication connection, which can be implemented as one or more communication buses or signal lines.

[0236] Through the above description of the embodiments, those skilled in the art can clearly understand that this application can be implemented by means of software plus necessary general-purpose hardware, or it can be implemented by special-purpose hardware including application-specific integrated circuits, special-purpose CPUs, special-purpose memory, special-purpose components, etc. Generally, any function performed by a computer program can be easily implemented by corresponding hardware, and the specific hardware structure used to implement the same function can also be diverse, such as analog circuits, digital circuits, or special-purpose circuits. However, for this application, software program implementation is more often the preferred implementation method. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a readable storage medium, such as a computer floppy disk, USB flash drive, mobile hard disk, ROM, RAM, magnetic disk, or optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, training equipment, or network device, etc.) to execute the methods described in the various embodiments of this application.

[0237] In the above embodiments, the implementation can be achieved, in whole or in part, through software, hardware, firmware, or any combination thereof. When implemented in software, it can be implemented, in whole or in part, in the form of a computer program product.

[0238] The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, all or part of the processes or functions described in the embodiments of this application are generated. The computer may be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions may be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions may be transmitted from one website, computer, training device, or data center to another website, computer, training device, or data center via wired (e.g., coaxial cable, fiber optic, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium may be any available medium that a computer can store or a data storage device such as a training device or data center that integrates one or more available media. The available media may be magnetic media (e.g., floppy disks, hard disks, magnetic tapes), optical media (e.g., DVDs), or semiconductor media (e.g., solid-state drives (SSDs)).

Claims

1. A data processing method, characterized in that, The method includes: Acquire first information and second information, wherein the first information includes the road structure of the driving scenario and the second information includes the category of driving behavior of the target vehicle in the driving scenario; Based on the first information and the second information, a feature representation is obtained through an encoder; Based on the feature representation, the information of the simulation scene of the driving scenario is obtained by generating a model. The information of the simulation scene includes the position and movement information of multiple objects, including the target vehicle, in the driving scenario.

2. The method according to claim 1, characterized in that, In the simulation scenario, the target vehicle interacts with other objects among the plurality of objects to perform the driving behavior of the aforementioned category.

3. The method according to claim 1 or 2, characterized in that, The driving behavior is classified into one of the following categories: Vehicle obstacle avoidance, lane keeping, lane changing, lane cutting, and merging into traffic.

4. The method according to any one of claims 1 to 3, characterized in that, The first information also includes the location information of at least one obstacle located in the driving scenario; the simulation scenario also includes the at least one obstacle.

5. The method according to any one of claims 1 to 4, characterized in that, The method further includes: Obtain the training effect of training the machine learning model based on the simulation scenario; The encoder or the generative model is updated with the goal of optimizing the training effect.

6. The method according to claim 4, characterized in that, The step of obtaining a feature representation through an encoder based on the first information and the second information includes: Based on the first information and the prompt containing the second information, the feature representation is obtained by the encoder.

7. The method according to claim 5 or 6, characterized in that, The encoder includes a first encoder and a second encoder; The step of obtaining feature representations based on the first information and the second information through an encoder includes: Based on the first information, the feature representation of the first information is obtained through the first encoder; Based on the second information, the feature representation of the second information is obtained through the second encoder; Updating the encoder includes: Update the second encoder.

8. The method according to any one of claims 1 to 7, characterized in that, The second encoder is a transformer-based neural network model.

9. The method according to any one of claims 1 to 8, characterized in that, The second information also includes the driving style of the target vehicle.

10. The method according to any one of claims 1 to 9, characterized in that, The method further includes: The generative model is updated based on the relationship between the movement information and the kinematic constraints.

11. The method according to any one of claims 1 to 10, characterized in that, The generative model is a diffusion model.

12. A data processing apparatus, characterized in that, The device includes: The acquisition module is used to acquire first information and second information, wherein the first information includes the road structure of the driving scenario and the second information includes the category of driving behavior of the target vehicle in the driving scenario; The processing module is used to obtain a feature representation by an encoder based on the first information and the second information; and to obtain a simulation scene of the driving scenario by generating a model based on the feature representation.

13. The apparatus according to claim 12, characterized in that, The simulation scenario includes movement information of multiple objects, including the target vehicle, in the driving scenario, and the target vehicle in the simulation scenario performs the driving behavior of the aforementioned category.

14. The apparatus according to claim 12 or 13, characterized in that, The driving behavior is classified into one of the following categories: Vehicle obstacle avoidance, lane keeping, lane changing, lane cutting, and merging into traffic.

15. The apparatus according to any one of claims 12 to 14, characterized in that, The processing module is further configured to: Obtain the training effect of training the machine learning model based on the simulation scenario; The encoder is updated with the goal of optimizing the training effect.

16. The apparatus according to claim 15, characterized in that, The second information specifically refers to the feature representation obtained through the encoder, including: Based on the first information and the prompt containing the second information, the feature representation is obtained by the encoder.

17. The apparatus according to claim 15 or 16, characterized in that, The encoder includes a first encoder and a second encoder; The processing module is specifically used for: Based on the first information, the feature representation of the first information is obtained through the first encoder; Based on the second information, the feature representation of the second information is obtained through the second encoder; Update the second encoder.

18. The apparatus according to any one of claims 12 to 17, characterized in that, The second encoder is a transformer-based neural network model.

19. The apparatus according to any one of claims 12 to 18, characterized in that, The second information also includes the driving style of the target vehicle.

20. The apparatus according to any one of claims 12 to 19, characterized in that, The processing module is further configured to: The generative model is updated based on the relationship between the movement information and the kinematic constraints.

21. The apparatus according to any one of claims 12 to 19, characterized in that, The generative model is a diffusion model.

22. A computer storage medium, characterized in that, The computer storage medium stores one or more instructions, which, when executed by one or more computers, cause the one or more computers to perform the operation of the method according to any one of claims 1 to 11.

23. A computer program product, characterized in that, Includes computer-readable instructions that, when executed on a computer device, cause the computer device to perform the method as described in any one of claims 1 to 11.

24. A system comprising at least one processor and at least one memory; the processor and the memory are connected via a communication bus and communicate with each other. The at least one memory is used to store code; The at least one processor is used to execute the code to perform the method as described in any one of claims 1 to 11.

25. A chip, characterized in that, It includes at least one processing unit and an interface circuit, the interface circuit being used to provide program instructions or data to the at least one processing unit, the at least one processing unit being used to execute the program instructions to implement the method of any one of claims 1 to 11.

Images