CN116450100A - Equipment development method and system based on generation type artificial intelligent model - Google Patents
Equipment development method and system based on generation type artificial intelligent model Download PDFInfo
- Publication number
- CN116450100A CN116450100A CN202310482739.XA CN202310482739A CN116450100A CN 116450100 A CN116450100 A CN 116450100A CN 202310482739 A CN202310482739 A CN 202310482739A CN 116450100 A CN116450100 A CN 116450100A
- Authority
- CN
- China
- Prior art keywords
- model
- generated artificial
- development
- artificial intelligent
- user
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000011161 development Methods 0.000 title claims abstract description 94
- 238000000034 method Methods 0.000 title claims abstract description 57
- 238000013528 artificial neural network Methods 0.000 claims abstract description 4
- 238000012549 training Methods 0.000 claims description 42
- 238000013473 artificial intelligence Methods 0.000 claims description 39
- 238000013461 design Methods 0.000 claims description 28
- 230000006870 function Effects 0.000 claims description 15
- 238000012986 modification Methods 0.000 claims description 10
- 230000004048 modification Effects 0.000 claims description 10
- 230000002452 interceptive effect Effects 0.000 claims description 6
- 238000012545 processing Methods 0.000 claims description 6
- 238000005728 strengthening Methods 0.000 claims description 5
- 238000004891 communication Methods 0.000 claims description 4
- 230000010354 integration Effects 0.000 claims description 4
- 239000000463 material Substances 0.000 claims description 3
- 238000003062 neural network model Methods 0.000 claims description 3
- 238000013526 transfer learning Methods 0.000 claims description 3
- 230000002093 peripheral effect Effects 0.000 claims description 2
- 238000012360 testing method Methods 0.000 abstract description 38
- 230000018109 developmental process Effects 0.000 description 59
- 230000006399 behavior Effects 0.000 description 33
- 238000004088 simulation Methods 0.000 description 21
- 230000015654 memory Effects 0.000 description 14
- 230000008676 import Effects 0.000 description 9
- 238000010586 diagram Methods 0.000 description 7
- 230000000007 visual effect Effects 0.000 description 7
- 230000003993 interaction Effects 0.000 description 6
- 230000008569 process Effects 0.000 description 6
- 230000005291 magnetic effect Effects 0.000 description 5
- 238000004590 computer program Methods 0.000 description 4
- 238000005516 engineering process Methods 0.000 description 3
- 238000013507 mapping Methods 0.000 description 3
- 230000006978 adaptation Effects 0.000 description 2
- 230000005540 biological transmission Effects 0.000 description 2
- 230000007547 defect Effects 0.000 description 2
- 238000001514 detection method Methods 0.000 description 2
- 239000012636 effector Substances 0.000 description 2
- 238000011156 evaluation Methods 0.000 description 2
- 230000003287 optical effect Effects 0.000 description 2
- 230000011218 segmentation Effects 0.000 description 2
- 230000033772 system development Effects 0.000 description 2
- 238000012795 verification Methods 0.000 description 2
- 241000282414 Homo sapiens Species 0.000 description 1
- 206010063385 Intellectualisation Diseases 0.000 description 1
- 239000008186 active pharmaceutical agent Substances 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000001413 cellular effect Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 238000007667 floating Methods 0.000 description 1
- 238000011423 initialization method Methods 0.000 description 1
- 230000004044 response Effects 0.000 description 1
- 230000003068 static effect Effects 0.000 description 1
- 230000009466 transformation Effects 0.000 description 1
- 238000012800 visualization Methods 0.000 description 1
- 230000003936 working memory Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F8/00—Arrangements for software engineering
- G06F8/20—Software design
- G06F8/24—Object-oriented
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/20—Natural language analysis
- G06F40/205—Parsing
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F8/00—Arrangements for software engineering
- G06F8/30—Creation or generation of source code
- G06F8/31—Programming languages or programming paradigms
- G06F8/315—Object-oriented languages
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F8/00—Arrangements for software engineering
- G06F8/30—Creation or generation of source code
- G06F8/33—Intelligent editors
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F8/00—Arrangements for software engineering
- G06F8/40—Transformation of program code
- G06F8/41—Compilation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- General Engineering & Computer Science (AREA)
- Software Systems (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- General Health & Medical Sciences (AREA)
- Computational Linguistics (AREA)
- Computing Systems (AREA)
- Artificial Intelligence (AREA)
- Health & Medical Sciences (AREA)
- Data Mining & Analysis (AREA)
- Biophysics (AREA)
- Biomedical Technology (AREA)
- Evolutionary Computation (AREA)
- Life Sciences & Earth Sciences (AREA)
- Molecular Biology (AREA)
- Mathematical Physics (AREA)
- Audiology, Speech & Language Pathology (AREA)
- Stored Programmes (AREA)
Abstract
The invention discloses a device development method and a system based on a generated artificial intelligent model, which relate to the technical fields of automatic device development, test device development, robot development, intelligent home development and device development of Internet of things, and comprise a generated artificial intelligent model based on a neural network, wherein the generated artificial intelligent model has natural language understanding and generating capability through learning of massive language sample data; and the device development is realized by utilizing the generated artificial intelligent model; the invention utilizes the natural language understanding capability and the formatted text generating capability of the generated artificial intelligent model, can generate and modify corresponding equipment configuration files or program language codes according to the natural language input by a user, and can describe and display the equipment configuration files in a flexible and accurate way; the method and the system can effectively solve the problems of the prior art that the method for developing test equipment by using natural language is lack, the equipment development has high requirements on engineering capability of development personnel, low development efficiency and the like.
Description
Technical Field
The invention relates to the technical fields of automatic equipment development, test equipment development, robot development, intelligent home development, internet of things development and the like, in particular to a method and a system for carrying out equipment development by using a generated artificial intelligent model.
Background
The existing equipment development method mainly comprises the steps that engineering personnel integrate various hardware modules into equipment or devices through designing computer software through understanding the system development requirements. In order to facilitate the development of equipment by engineering personnel, software such as Simulink, dSpace and the like is designed in a modularized manner by software and hardware, and low-code equipment is designed and developed in a visual connection mode or a configuration form filling mode and the like, so that the requirement on the development capability of the program code of the equipment developer is reduced, but the existing equipment development method still has the following defects:
(1) The equipment development has high requirements on engineering experience of personnel, and although part of tool software provides a technical means for carrying out equipment development in a low-code mode, the development personnel still need to have deeper understanding on the equipment composition principle, the application requirements are manually converted into engineering requirements, and a method and tools for directly converting the application requirements described by natural language into engineering configuration files are lacked.
(2) The equipment development lacks intelligent and automatic support, the complex system information flow design connection, interface configuration and other works are complex, the configuration workload is extremely large by means of manual modes such as visualization, errors are easy to occur, and debugging and verification are difficult.
(3) The equipment development is difficult to realize the balance of generalization and individuation, and because equipment developers cannot fully know the application requirements of equipment users, the existing equipment modules and configuration files cannot be fully utilized to realize the universalization design of the equipment, and the individuation use requirements of the users for different scenes and functions cannot be met.
The generated artificial intelligent model is a large-scale neural network model, is a general intelligent model with natural language understanding capability, and is represented by GPT, so that human beings and computers are enabled to be realized through natural language interaction. The generated artificial intelligent model can learn rich language knowledge and logic reasoning capability from massive text data, understand the user requirements of natural language expression and generate proper response according to the context. The generated artificial intelligent model can adapt to data of different fields or tasks through fine-tuning (transfer learning), and can utilize a description document with a fixed format designed by engineering personnel as a training sample, such as a configuration file or program code, and the like, so that the efficiency of the model in generating the document is greatly improved.
The invention aims to provide a device design method and a system by utilizing a related technology of a generated artificial intelligent model, which can overcome the defects, improve the development efficiency and quality of the device and realize the generalization and individuation of the device.
Disclosure of Invention
The invention relates to a device development method and a system based on a generated artificial intelligent model, which utilize the natural language understanding capability and formatted text generating capability of the generated artificial intelligent model to carry out device design, thereby improving the efficiency and quality of device development. The technical scheme of the invention is as follows:
the invention adopts a generating artificial intelligent model based on a neural network, and the generating artificial intelligent model has natural language understanding and generating capability through learning of massive language sample data; and the device development is realized by using the generated artificial intelligence model.
Further, the developing steps of the generating artificial intelligence model implementation equipment are as follows:
s1: in the initial design stage, a specific engineering field is selected, and a standardized program framework template and a configuration file format are designed;
s2: in the design stage, collecting, sorting and compiling equipment development samples under a typical application scene in the engineering field;
s3: at the final design stage, training or fine-tuning the generated artificial intelligent model by using the equipment development sample, and strengthening the equipment development related capability of the training generated artificial intelligent model;
s4: and in the using stage of the user, the system receives a task request sent by the user, and generates and modifies configuration files or program language codes sent by the user, displays information flows and control flows, consults information and compiles a plurality of running control requests in an interactive mode.
Further, the generated artificial intelligence model is a neural network model, and functions which can be realized by large-scale data set training comprise several of identification, summarization, prediction, text generation and code generation.
Further, in the design stage, equipment development samples under the typical application scene of the engineering field are collected, arranged and compiled, and the equipment development samples comprise the following characteristics:
a1: information flow and control flow configuration requirements of equipment in the engineering field under a typical application scene are collected, arranged and compiled, and corresponding configuration files or programming language codes are used for forming equipment configuration samples;
a2: optionally, collecting, sorting and compiling a plurality of interface document samples including a software function module interface document, a communication message module interface document, a hardware module interface document and a peripheral module interface in the engineering field;
a3: optionally, the display style of various configuration files or program language codes in the engineering field, which is convenient for the user to check, can be designed, and a plurality of formats including characters, graphics, tables, sounds, images and videos can be adopted to form a configuration display sample together with the configuration files or program language codes.
Further, at the design end stage, the device development sample is utilized to train or fine tune the generated artificial intelligent model, and the device development correlation capability of the generated artificial intelligent model is enhanced, which comprises the following characteristics:
b1: the generated artificial intelligent model is a large-scale pre-training model, general knowledge and capability are learned from massive data, and different downstream tasks are adapted through fine tuning or transfer learning;
b2: the training or fine tuning of the generated artificial intelligent model is to train or fine tune the generated artificial intelligent model by using samples in a device development sample library as training data, so that the generated artificial intelligent model can complete a plurality of tasks including configuration file or programming language code generation and modification, information flow and control flow display, information consultation and compiling operation control according to user operation instructions.
Further, in the user use stage, the system receives a task request proposed by a user, and interactively completes a plurality of types of configuration files or program language code generation and modification, information flow and control flow display and information consultation proposed by the user, including the following steps:
c1: the generated artificial intelligent model generates a plurality of corresponding configuration files, program language codes, information flows and control flow display and consultation replies according to the user request;
c2: optionally, the system receives feedback comments of the user on a plurality of contents in the display of the configuration file, the programming language code, the information flow and the control flow and the consultation reply, which are generated by the generated artificial intelligence model in the substep C1;
and C3: optionally, the generated artificial intelligent model correspondingly adjusts the output result according to the user feedback opinion and displays the output result to the user again;
and C4: optionally, repeating sub-steps C2 and C3 until the user is satisfied with or terminates the task including several of the configuration file, the programming language code, the information flow and control flow presentation, the advisory reply;
c5: storing a plurality of configuration files and programming language codes so as to facilitate compiling and running of equipment software;
c6: optionally, compiling and deploying the program language code.
Further, the system comprises a device development sample library which is responsible for storing and managing a plurality of sample materials in the device configuration sample, the interface document sample and the configuration display sample;
model training module: training or fine-tuning the generated artificial intelligent model to enable the generated artificial intelligent model to have equipment development related capability;
generating an artificial intelligence model: the system is responsible for analyzing and processing a plurality of tasks including configuration files or program code generation, configuration files or program code modification, configuration display and information consultation, which are proposed by a user, and giving required replies;
and a system integration module: the configuration file generated by integrating and deploying the generated artificial intelligence model is responsible for compiling and deploying the programming language code;
and a system input module: the method is in charge of receiving a plurality of inputs of a user in a mode of containing words, sounds, pictures and videos, and converting the inputs into a format which can be processed by the generated artificial intelligence model;
and a system output module: and the content generated by the generated artificial intelligent model is displayed to the user in a form comprising a plurality of text output, picture output, audio output and video output.
Further, the sample formats in the device development sample library comprise a plurality of words, program language codes, charts, pictures, models, drawings, sounds and videos.
Further, the model training module may comprise any one or more training algorithms for training the generative artificial intelligence model.
Further, the generated artificial intelligence model can be obtained by adopting an existing pre-training model and utilizing the device development sample library for fine adjustment or utilizing sample data comprising the device development sample library for training.
Compared with the prior art, the invention has the following beneficial effects:
(1) The invention utilizes the powerful language understanding capability and formatted text generating capability of the generated artificial intelligent model to realize the intellectualization and generalization of equipment development, and can quickly generate proper configuration files or program language codes according to the application requirements of different natural language descriptions, thereby improving the system development efficiency of the equipment;
(2) By standardizing the information flow and control flow design among the equipment modules, the invention simplifies the complexity of equipment development, is convenient for debugging and verification, reduces the degree of dependence of personnel experience and improves the system design quality;
(3) According to the invention, the sample library is developed through establishing equipment, and the generated artificial intelligent model is trained or fine-tuned, so that the model has intelligence and self-adaptability, and an optimal or most suitable equipment design scheme can be generated or adjusted according to different requirements and scenes;
(4) The invention provides a universal equipment design tool, and a user can finish the construction and transformation of equipment according to own requirements without depending on professional staff.
Drawings
FIG. 1 is a flow diagram of a method for developing a device based on a generated artificial intelligence model in accordance with the present invention;
FIG. 2 is a schematic diagram of an exemplary information flow diagram of a device development method based on a generative artificial intelligence model in accordance with the present invention;
FIG. 3 is a block diagram of an equipment development system based on a generative artificial intelligence model in accordance with the present invention;
Detailed Description
In order to make the objects, technical solutions and advantages of the embodiments of the present application more clear, the technical solutions of the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application, and it is apparent that the described embodiments are some embodiments of the present application, but not all embodiments. Based on the embodiments herein, all other embodiments that a person of ordinary skill in the art could achieve without inventive effort are within the scope of the present application.
As used herein, the singular forms "a", "an", "the" and "the" are intended to include the plural forms as well, unless expressly stated otherwise, as understood by those skilled in the art. It will be further understood that the terms "comprises" and/or "comprising," when used in this specification, specify the presence of stated features, integers, steps, operations, elements, and/or components, but do not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and/or groups thereof.
Examples
Referring to fig. 1, the invention provides a device development method based on a generated artificial intelligent model, which is suitable for tasks of configuration file generation, modification, display and consultation in the development process of semi-physical simulation devices.
The equipment development method based on the generated artificial intelligent model comprises a generated artificial intelligent model based on a neural network, wherein the generated artificial intelligent model has natural language understanding and generating capability through learning of massive language sample data; and the device development is realized by using the generated artificial intelligence model.
In a preferred embodiment, a GPT-3-S generated artificial intelligent model is taken as an example for illustration in the embodiment, the model has 1.25 hundred million model parameters, and the model has understanding and generating capability of Chinese and English natural language through pre-training of massive language sample data; the task of developing semi-physical simulation equipment is carried out by utilizing the generated artificial intelligent model, and the method comprises the following specific steps:
s1: in the initial design stage, a specific engineering field is selected, and a standardized program framework template and a configuration file format are designed;
in a preferred embodiment, the field of development of the semi-physical simulation equipment is specifically described, and the terminal simulator equipment is to be developed, so that the behavior of the terminal CAN be simulated, and the terminal CAN be communicated with the master controller in a CAN bus and RS422 communication mode. The terminal adopts a modularized design and is divided into a behavior simulation module, a message module and a hardware module, wherein the behavior simulation module is responsible for simulating the behavior of the terminal, the message module is responsible for grouping and unpacking message data, and the hardware module is responsible for information transmission of CAN and RS422 buses. The behavior simulation module adopts an FMI interface standard, the message module can load a standard configuration file to realize custom package and unpack, the hardware module calls related hardware API to realize communication, and the behavior simulation module and the message module realize data interaction according to the standard information flow configuration file.
In the initial stage of design, simulator equipment software and hardware are subjected to modularized design, standard interfaces of all modules are defined, and a whole simulator software framework is built, wherein a typical software framework pseudo code is shown as follows.
Main program framework:
# import correlation module
import FMI
import Message
import Hardware
# definition terminal simulator class
class TerminalSimulator:
# initialization method, incoming behavior simulation file, message configuration file, information flow configuration file, hardware type and address
def __init__(self, behavior_file, message_file, infoflow_file, hardware_type, hardware_address):
Creating behavior simulation module object and loading behavior simulation file
self.behavior = FMI.FMU(behavior_file)
Creating message module object and loading message configuration file
self.message = Message.Message(message_file)
... ...
def load_infoflow(self, infoflow_file):
# open information stream configuration file, read content
with open(infoflow_file) as f:
content = f.read()
... ...
Dictionary of# return data interaction rules
return infoflow_dict
Method for analyzing information flow configuration file content, transmitting content character string and returning dictionary of data interaction rule
def parse_infoflow(self, content):
Specific implementation details are omitted, and an example dictionary is assumed to be returned
return {"behavior_output_1": "message_input_1", "message_output_2": "behavior_input_2"}
# main circulation method for realizing function of terminal simulator
def main_loop(self):
The following steps are performed in a # loop until the stop condition is satisfied
while True:
The do_step method of the behavior simulation module is called for # to execute one-step simulation, and a simulation result is returned
behavior_result = self.behavior.do_step()
Mapping the output of the behavior simulation module to the input of the message module according to the rule of the information flow configuration file, and assigning values to the message input variables
... ...
The set_input method of the behavior simulation module is called in a # mode, simulation input is set according to behavior input variables, and whether success is achieved or not is returned
success = self.behavior.set_input(behavior_input)
If the setting input fails, # print error information and exit the loop
if not success:
print("Error: invalid command data")
break
Method for mapping output of one module to input of another module according to rule of information flow configuration file, input-output data, output type and input type, return input data
def map_data(self, output_data, output_type, input_type):
# create an empty input data dictionary
input_data = {}
Rule dictionary for # traversing information flow configuration file
... ...
Return to input data dictionary
return input_data
Establishing a terminal simulator object and transmitting related parameters
terminal_simulator = TerminalSimulator("behavior.fmu", "message.xml", "infoflow.xml", "CAN", "0x1234")
The main_loop method of the terminal simulator object is called in a# mode, and the terminal simulator program starts to run
terminal_simulator.main_loop()
Defining information flow mapping relation between the behavior simulation module and the message module, and adopting a typical information flow configuration file format defined by XML format as follows:
<?xml version="1.0" encoding="UTF-8"?>
<infoflow>
the < | -information flow configuration file is used for defining data interaction rules between the behavior simulation module and the message module
< |— each rule element represents a rule containing five attributes: source_module, source_port, target_module, target_port, info-
The source module attribute represents the source module of the data and the target module attribute represents the destination module of the data
The < | - -source_port attribute represents the source port of the data and the target_port attribute represents the destination port of the data- >
The < | - -info attribute represents other related information, such as data type, unit, range, etc. - >
The values of the source_module and target_module attributes should be either behavir or message- >
Values of the < | - -source_port and target_port attributes should be consistent with port names defined in behavior simulation files and message configuration files- - - -
The first rule below indicates, for example, that the output port behavior_output_1 of the behavior simulation module is mapped to the input port message_input_1 of the message module, the data type is int, in m/s, in the range of 0-100- >
<rule source_module="XX" source_port="XX" target_module="XX" target_port="XX" info="type=XX, unit=XX, range=XX"/>
</infoflow>
The standard configuration file format for packet and unpacking by the definition message module is as follows, and a typical message configuration file format defined by XML format is as follows:
<?xml version="1.0" encoding="UTF-8"?>
<message>
the </I > -message configuration file is used for defining grouping and unpacking rules of message data- >
< |— each data element represents one data, containing five attributes: name, type, position, endian, info- >
The < | - -name attribute represents the name of the data, and the type attribute represents the type of the data- -
The < | -position attribute represents the position of the data in the message, counting from 0, and the endian attribute represents the size end of the data >
The < | - -info attribute represents other related information, such as units, ranges, enumerated values, etc. - >
The value of the < | — name attribute should be consistent with the port name defined in the information flow profile- >
The value of the < | -type attribute should be the basic type or custom type of int, float, pool, string, etc. -)
The value of the < | -position attribute should be an integer and the value of the endian attribute should be big or lite- >
The first data element below represents, for example, a data named message_input_1, of type int, occupies the 0-3 rd byte in the message, is big at the size end, is m/s, and ranges from 0-100- >
<data name="XX" type="XX" position="XX" endian="XX" info="unit=XX, range=XX"/>
</message>
S2: in the design stage, collecting, sorting and compiling equipment development samples under a typical application scene in the engineering field;
in a preferred embodiment, the information flow and control flow configuration requirements of the terminal simulator equipment in a typical application scene and corresponding configuration files are collected, arranged and compiled to form equipment configuration samples; a typical device configuration sample is as follows:
configuration requirements:
the output port behavir_output_1 of the behavior simulation module is mapped to the input port message_input_1 of the message module, the data type is int, the unit is m/s, and the range is 0-100. The first data element in the message configuration file represents a data named message_input_1, the type is int, the 0-3 rd byte in the message is occupied, the large end represents the unit is m/s, and the range is 0-100.
Corresponding information flow configuration files:
<?xml version="1.0" encoding="UTF-8"?>
<infoflow>
<rule source_module="behavior" source_port="behavior_output_1" target_module="message" target_port="message_input_1" info="type=int, unit=m/s, range=0-100"/>
</infoflow>
corresponding message configuration files:
<?xml version="1.0" encoding="UTF-8"?>
<message>
<data name="message_input_1" type="int" position="0" endian="big" info="unit=m/s, range=0-100"/>
</message>
the configuration file display style which is convenient for the user to check is designed, and a typical configuration display sample is shown as follows in a form mode:
information flow configuration file:
<?xml version="1.0" encoding="UTF-8"?>
<infoflow>
<rule source_module="behavior" source_port="behavior_output_1" target_module="message" target_port="message_input_1" info="type=int, unit=m/s, range=0-100"/>
</infoflow>
and displaying the corresponding configuration file:
| source_module | source_port | target_module | target_port | info |
| behavior | behavior_output_1 | message | message_input_1 | type=int, unit=m/s, range=0-100 |
s3: at the final design stage, training or fine-tuning the generated artificial intelligent model by using the equipment development sample, and strengthening the equipment development related capability of the training generated artificial intelligent model;
in a preferred embodiment, the GPT-3-S pre-training model is specifically tuned using python, and the sample set is a sample including a device development sample and a configuration display sample.
The data set is loaded and preprocessed, such as by loading the sample data set using a dataset library and encoding the text using a GPT-3 word segmentation engine provided by a transformers library.
A model configuration and optimizer is defined, such as using an AutoConfig class provided by the transformers library to load the configuration of GPT-3, and an AdamW optimizer to update model parameters.
Training parameters and evaluation indicators are defined, training parameters are set using the trainingargues class provided by the transformers library, and confusion is loaded using the load_metric function provided by the datasets library as an evaluation indicator. The parameters were set as follows:
batch_size: 8
learning_rate: 5e-5
num_train_epochs: 3
gradient_accumulation_steps: 8
max_grad_norm: 1.0
s4: and in the using stage of the user, the system receives a task request sent by the user, and generates and modifies configuration files or program language codes sent by the user, displays information flows and control flows, consults information and compiles a plurality of running control requests in an interactive mode.
In a preferred embodiment, it is specifically described that the user puts out a demand to the system in text input form, "output_1 and output_2 parameters of the behavir module are transmitted through the CAN bus, output_2 and output_3 parameters are transmitted through the RS422 bus, each parameter occupies 4 bytes, the first parameter is integer, the range is 0-100, and the second and third parameters are both floating point type.
The large language model generates information flow configuration files according to user requests and displays the information flow configuration files in a form so as to be convenient for the user to confirm, and a typical information flow configuration is shown as follows:
| source_module | source_port | target_module | target_port | info |
| Behavior | output_1 | Message1 | input_1 | type=int, unit=m/s, range=[0,100] |
| Behavior | output_2 | Message1 | input_2 | type=float, unit=N/A, range=N/A |
| Behavior | output_2 | Message2 | input_1 | type=int, unit=m/s, range=N/A |
| Behavior | output_3 | Message2 | input_2 | type=float, unit=N/A, range=N/A |
| Message1 | output | CAN | input | type=pack unit=N/A, range=N/A |
| Message2 | output | RS422 | input | type=pack unit=N/A, range=N/A |
or in the form of an information flow diagram, a typical information flow diagram is shown in fig. 2:
the user makes an interactive request "the second parameter range of the behavir module is set to-1 to 1" by confirming that the behavir_output_2 parameter is found to have no parameter range set. The generated artificial intelligence model regenerates the information flow configuration file and the configuration display and displays the information flow configuration file and the configuration display to a user as follows:
| source_module | source_port | target_module | target_port | info |
| Behavior | output_1 | Message1 | input_1 | type=int, unit=m/s, range=[0,100] |
| Behavior | output_2 | Message1 | input_2 | type=float, unit=N/A, range=[-1,1] |
| Behavior | output_2 | Message2 | input_1 | type=int, unit=m/s, range=[-1,1] |
| Behavior | output_3 | Message2 | input_2 | type=float, unit=N/A, range=N/A |
| Message1 | output | CAN | input | type=pack unit=N/A, range=N/A |
| Message2 | output | RS422 | input | type=pack unit=N/A, range=N/A |
the user confirms that the configuration is correct, replies "confirm configuration" to the system, and can generate a new device. The system stores the information flow configuration file and the message configuration file confirmed by the user. And restarting the equipment by the system, loading the newly generated information flow configuration file and the newly generated message configuration file by the program main framework, and executing the information flow logic newly defined by the user.
Example two
Referring to fig. 1, the invention provides a device development method based on a generated artificial intelligent model, which is suitable for program code generation tasks in the test device development process. The field of development of the testing equipment is selected, and the testing equipment for testing the grabbing function of the industrial robot is to be developed. The grabbing function refers to that the industrial robot to be detected picks up the object to be grabbed by utilizing the visual guidance and the grippers thereof and places the object to be grabbed at a specified position. The test equipment consists of a mechanical arm, a controller and a visual sensor, wherein the mechanical arm is responsible for placing the object to be grabbed at a certain position near the industrial robot, and the visual sensor is responsible for detecting whether the object to be grabbed is placed at a specified position. In the test process, the mechanical arm of the test system places the object to be grabbed at a certain position according to the setting of the test case, the industrial robot picks up the object to be grabbed and places the object to be grabbed at a specified position, and the test system judges whether the test is passed or not by detecting whether the industrial robot places the object to be grabbed at the specified position.
The model selection condition is the same as that of the first embodiment, and the task of developing test equipment is carried out by utilizing the generated artificial intelligent model, and the specific steps are as follows:
s1 the method comprises the following steps: in the initial design stage, a specific engineering field is selected, and a standardized program framework template and a configuration file format are designed;
in a preferred embodiment, the standard interface functions of the robotic arm and vision sensor are specifically designed based on the composition of the test system.
The mechanical arm consists of a controller and an arm, and the arm has six degrees of freedom and can move and rotate in a three-dimensional space. The robotic arm controller may receive the following instructions:
move (x, y, z) move the end effector of the arm to coordinates (x, y, z).
Rotate (rx, ry, rz) rotate the end effector of the arm by rx degrees, ry degrees and rz degrees.
Grasp ()' closing the jaws of the arm, gripping the object.
Release ()'s opening the jaws of the arm, releasing the object.
The vision sensor is composed of a camera and a depth sensor, and can capture images and depth information in a scene. The vision sensor may return the following data:
image-an RGB image representing color information in the scene.
Depth, a gray scale image representing depth information in the scene.
Mask-a binary image representing the result of the segmentation of objects in the scene.
Bbox-a list representing object bounding box coordinates in the scene.
A list representing object pose parameters in the scene.
S2: in the design stage, collecting, sorting and compiling equipment development samples under a typical application scene in the engineering field;
in a preferred embodiment, a typical test equipment development sample is specified as follows:
instructions to: the mechanical arm is used for placing the bottle on the chair and detecting whether the bottle exists on the chair
And (3) outputting:
module of # leading-in mechanical arm and vision sensor
import arm
import vision
# initializing mechanical arm and vision sensor
arm = arm.Arm()
vision = vision.Vision()
# acquisition of data of visual sensor
image = vision.get_image()
depth = vision.get_depth()
mask = vision.get_mask()
bbox = vision.get_bbox()
pose = vision.get_pose()
Move the robotic arm to the vial position #
arm.move(bottle_x, bottle_y, bottle_z)
# grabbing bottle
arm.grasp()
Target position on # positioning chair
chair_index = mask.index("chair")
chair_x, chair_y, chair_z = pose[chair_index]
target_x=chair_x+0.1# assume that the target position is 10 cm to the right of the center of the chair
target_y = chair_y
target_z=chair_z+0.5# assume that the target position is 50 cm above the chair surface
Moving the mechanical arm to the target position #
arm.move(target_x, target_y, target_z)
# release bottle
arm.release()
Test results of # detection
result=vision. Check_sensor (target_x, target_y, target_z) # assume that the vision sensor can detect whether or not there is an object at the specified position
if result == "bottle":
print ('test success')
else:
print ('test failure')
S3: at the final design stage, training or fine-tuning the generated artificial intelligent model by using the equipment development sample, and strengthening the equipment development related capability of the training generated artificial intelligent model;
in a preferred embodiment, the same embodiment as the first embodiment is specifically described.
S4: and in the using stage of the user, the system receives a task request sent by the user, and generates and modifies configuration files or program language codes sent by the user, displays information flows and control flows, consults information and compiles a plurality of running control requests in an interactive mode.
In a preferred embodiment, the user description task request is specifically described as follows: "please design test equipment, the pick-up function of the industrial robot can be tested according to a given set of test cases. The test set is for example as follows:
1) Test case one: the object is 1 meter away from the base, the azimuth angle is 0 degree, and the object is in the working range and the visual range of the industrial robot to be tested;
2) Test case two: the object is 1.5 m away from the base, the azimuth angle is 45 degrees, and the object is in the working range of the industrial robot to be tested but is not in the visible range;
3) Test case three: the object is 2 meters away from the base, the azimuth angle is 0 degree, and the object is not in the working range of the industrial robot to be tested;
the system generates the programming language code according to the user request as follows:
module for # importing mechanical arm of industrial robot to be tested and testing system
import robot
import test_arm
import vision
Mechanical arm for initializing industrial robot to be tested and test system
robot = robot.Robot()
test_arm = test_arm.Arm()
vision = vision.Vision()
# describes test cases, including different object positions
test_cases = [
{"distance": 1, "azimuth": 0}
{"distance": 1.5, "azimuth": 45}
{"distance": 2, "azimuth": 0}
]
# acquisition of data of visual sensor
image = vision.get_image()
depth = vision.get_depth()
mask = vision.get_mask()
bbox = vision.get_bbox()
pose = vision.get_pose()
# traverse all test cases
for i in range(len(test_cases)):
# calculating coordinates of an object from the object position
ball_distance = test_cases[i]["distance"]
ball_azimuth = test_cases[i]["azimuth"]
ball_x = ball_distance * math.cos(math.radians(ball_azimuth))
ball_y = ball_distance * math.sin(math.radians(ball_azimuth))
ball_z=0.1# assumes that the object is placed on the ground
The mechanical arm of the # control test system places the object at the corresponding position
test_arm.move(ball_x, ball_y, ball_z)
test_arm.release()
Position of # positioning object
ball_index = mask.index("ball")
ball_x, ball_y, ball_z = pose[ball_index]
# determining whether an object is within the working range and the visible range of the industrial robot under test
ball_distance = math.sqrt(ball_x**2 + ball_y**2 + ball_z**2)
ball_azimuth = math.degrees(math.atan2(ball_y, ball_x))
if ball_distance <= 1.5 and abs(ball_azimuth) <= 45:
# control the measured industrial robot to move to the object position
robot.move(ball_x, ball_y, ball_z)
# attempt to grasp an object
robot.grasp()
# detection of the grasp results
result = vision.check_grasp()
Record test data #
print (f "test data: object position of test case { i } is ({ ball_x }, { ball_y }, { ball_z }), and grabbing result is { result }"
# release object
robot.release()
Restoring the industrial robot to be tested to the initial position #, and
robot.move(0, 0, 0)
else:
record test data #
print (f "test data: the object position of the test case { i } is ({ ball_x }, { ball_y }, { ball_z }) and cannot be grasped")
The system integration module deploys the program code, compiles and executes the program code, and the test equipment controller controls the mechanical arm to place the small ball at the designated position and observes whether the tested industrial robot grabs the small cell to the designated area.
Example III
As shown in fig. 3, which is a structural diagram of a device development system based on a generated artificial intelligence model of the present invention, it includes:
device development sample library 11: is responsible for storing and managing a plurality of sample materials in the equipment configuration sample, the interface document sample and the configuration display sample; the sample formats in the device development sample library 11 include several of text, program language codes, charts, pictures, models, drawings, sounds and videos.
Model training module 12: is responsible for training or fine tuning the generated artificial intelligence model 13, so that the generated artificial intelligence model 13 has equipment development related capability; wherein the model training module 12 may comprise any one or more training algorithms for training the generative artificial intelligence model 13.
Generating an artificial intelligence model 13: the system is responsible for analyzing and processing a plurality of tasks including configuration files or program code generation, configuration files or program code modification, configuration display and information consultation, which are proposed by a user, and giving required replies; the generated artificial intelligence model 13 can be obtained by adopting the existing pre-training model and utilizing the device development sample library 11 to finely tune, or by utilizing sample data comprising the device development sample library 11 to train.
System integration module 14: the configuration file generated by integrating and deploying the generated artificial intelligence model is responsible for compiling and deploying the programming language code;
system input module 15: the method is in charge of receiving a plurality of inputs of a user in a mode of containing words, sounds, pictures and videos, and converting the inputs into a format which can be processed by the generated artificial intelligence model;
system output module 16: and the content generated by the generated artificial intelligent model is displayed to the user in a form comprising a plurality of text output, picture output, audio output and video output.
The foregoing is only a preferred embodiment of the present invention, and the scope of the present invention is not limited to the foregoing examples, but all technical solutions belonging to the concept of the present invention are within the scope of the present invention. It should be noted that modifications and adaptations to the present invention may occur to one skilled in the art without departing from the principles of the present invention and are intended to be within the scope of the present invention.
The technical features of the above-described embodiments may be arbitrarily combined, and all possible combinations of the technical features in the above-described embodiments are not described for brevity of description, however, as long as there is no contradiction between the combinations of the technical features, they should be considered as the scope of the description.
It should be noted that the above embodiments can be freely combined as needed. The foregoing is merely a preferred embodiment of the present invention and it should be noted that modifications and adaptations to those skilled in the art may be made without departing from the principles of the present invention, which are intended to be comprehended within the scope of the present invention.
The software program of the present invention may be executed by a processor to perform the steps or functions described above. Likewise, the software programs of the present invention (including associated data structures) may be stored on a computer readable recording medium, such as RAM memory, magnetic or optical drive or diskette and the like. In addition, some of the steps or functions of the present invention may be implemented in hardware, for example, as circuitry that cooperates with the processor to perform various functions or steps. The methods disclosed in the embodiments shown in the embodiments of the present specification may be applied to a processor or implemented by a processor. The processor may be an integrated circuit chip having signal processing capabilities. In implementation, the steps of the above method may be performed by integrated logic circuits of hardware in a processor or by instructions in the form of software. The processor may be a general purpose processor including a central processing unit (Central Processing UnitCPU) network processor (Net work ProcessorP), a digital signal processor (Digita Signal ProcessorDSP), an application specific integrated circuit (Application Specific Integrated CircuitASIC) Field programmable gate array (Field-Programmable Gate Array FPGA) or other programmable logic device, discrete gate or body tube logic device, discrete hardware components, etc. The various methods, steps and logic blocks disclosed in the embodiments of this specification may be implemented or performed. A general purpose processor may be a microprocessor or the processor may be any conventional processor or the like. The steps of a method disclosed in connection with the embodiments of the present specification may be embodied directly in hardware, in a decoded processor, or in a combination of hardware and software modules in a decoded processor. The software modules may be located in a random access memory, flash memory, read only memory, programmable read only memory, or electrically erasable programmable memory, registers, etc. as well known in the art. The storage medium is located in a memory, and the processor reads the information in the memory and, in combination with its hardware, performs the steps of the above method.
Any of the modules of the system can be independently or integrally operated on an electronic device, and the electronic device can be any electronic product which can perform man-machine interaction with a user, such as a personal computer, a tablet computer, a smart phone, a personal digital assistant (Personal Digital AssistantPDA), a game console, an interactive network television (Internet Protocol TelevisionIPTV), an intelligent wearable device and the like. The electronic device may also include a network device and/or a user device. Wherein the network device includes, but is not limited to, a single network server, a server group composed of a plurality of network servers, or a Cloud based Cloud Computing (Cloud Computing) composed of a large number of hosts or network servers. The network in which the electronic device is located includes, but is not limited to, the internet, a wide area network, a metropolitan area network, a local area network, a virtual private network (Virtual Private NetworkVPN), and the like.
The system, apparatus, module or unit set forth in the above embodiments may be implemented in particular by a computer chip or entity or by a product having some function. One typical implementation is a computer. In particular, the computer may be, for example, a personal computer, a laptop computer, a cellular telephone, a camera phone, a smart phone, a personal digital assistant, a media player, a navigation device, an email device, a game console, a tablet computer, a wearable device, or a combination of any of these devices. Computer readable media, including both non-transitory and non-transitory, removable and non-removable media, may implement information storage by any method or technology. The information may be computer readable instructions, data structures, modules of a program, or other data. Examples of storage media for a computer include, but are not limited to, phase change memory (PRAM), static Random Access Memory (SRAM), dynamic Random Access Memory (DRAM), other types of Random Access Memory (RAM), read Only Memory (ROM), electrically Erasable Programmable Read Only Memory (EEPROM), flash memory or other memory technology, compact disc read only memory (CD-ROM), digital Versatile Discs (DVD) or other optical storage, magnetic cassettes, magnetic tape magnetic disk storage or other magnetic storage devices, or any other non-transmission medium, which can be used to store information that can be accessed by a computing device. Computer-readable media, as defined herein, does not include transitory computer-readable media (transshipment) such as modulated data signals and carrier waves. It should also be noted that the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one … …" does not exclude the presence of other like elements in a process, method, article or apparatus that comprises the element.
Furthermore, portions of the present invention may be implemented as a computer program product, such as computer program instructions, which when executed by a computer, may invoke or provide methods and/or techniques in accordance with the present invention by way of operation of the computer. Program instructions for carrying out the methods of the present invention may be stored on a fixed or removable recording medium and/or transmitted over a data stream on a broadcast or other signal bearing medium and/or stored in a working memory of a computer device operating in accordance with the program instructions. An embodiment according to the invention comprises an apparatus comprising a memory for storing computer program instructions and a processor for executing the program instructions, wherein the computer program instructions, when executed by the processor, trigger the apparatus to operate a method and/or a solution according to the embodiments of the invention as described above.
Claims (10)
1. The equipment development method based on the generated artificial intelligent model comprises the steps of generating the generated artificial intelligent model based on a neural network and utilizing the generated artificial intelligent model to realize equipment development.
2. The method for developing equipment based on the generated artificial intelligence model according to claim 1, wherein the method comprises the following steps: the method for developing the generated artificial intelligent model comprises the following steps of:
s1: in the initial design stage, a specific engineering field is selected, and a standardized program framework template and a configuration file format are designed;
s2: in the design stage, collecting, sorting and compiling equipment development samples under a typical application scene in the engineering field;
s3: at the final design stage, training or fine-tuning the generated artificial intelligent model by using the equipment development sample, and strengthening the equipment development related capability of the training generated artificial intelligent model;
s4: and in the using stage of the user, the system receives a task request sent by the user, and generates and modifies configuration files or program language codes sent by the user, displays information flows and control flows, consults information and compiles a plurality of running control requests in an interactive mode.
3. The method for developing equipment based on the generated artificial intelligence model according to claim 1, wherein the method comprises the following steps: the generated artificial intelligence model is a neural network model, and functions which can be realized by large-scale data set training comprise a plurality of identification, summarization, prediction, text generation and code generation.
4. The device development method based on the generated artificial intelligence model according to claim 2, wherein: in the design stage, equipment development samples under the typical application scene of the engineering field are collected, arranged and compiled, and the equipment development samples comprise the following characteristics:
a1: information flow and control flow configuration requirements of equipment in the engineering field under a typical application scene are collected, arranged and compiled, and corresponding configuration files or programming language codes are used for forming equipment configuration samples;
a2: optionally, collecting, sorting and compiling a plurality of interface document samples including a software function module interface document, a communication message module interface document, a hardware module interface document and a peripheral module interface in the engineering field;
a3: optionally, the display style of various configuration files or program language codes in the engineering field, which is convenient for the user to check, can be designed, and a plurality of formats including characters, graphics, tables, sounds, images and videos can be adopted to form a configuration display sample together with the configuration files or program language codes.
5. The device development method based on the generated artificial intelligence model according to claim 2, wherein: at the end of design, training or fine-tuning the generated artificial intelligent model by using the equipment development sample, and strengthening the equipment development related capability of the training generated artificial intelligent model, wherein the equipment development related capability comprises the following characteristics:
b1: the generated artificial intelligent model is a large-scale pre-training model, general knowledge and capability are learned from massive data, and different downstream tasks are adapted through fine tuning or transfer learning;
b2: the training or fine tuning of the generated artificial intelligent model is to train or fine tune the generated artificial intelligent model by using samples in a device development sample library as training data, so that the generated artificial intelligent model can complete a plurality of tasks including configuration file or programming language code generation and modification, information flow and control flow display, information consultation and compiling operation control according to user operation instructions.
6. The device development method based on the generated artificial intelligence model according to claim 2, wherein: in the using stage of the user, the system receives the task request proposed by the user, and interactively completes a plurality of types of generation and modification of configuration files or programming language codes, information flow and control flow display and information consultation proposed by the user, and the method comprises the following steps:
c1: the generated artificial intelligent model generates a plurality of corresponding configuration files, program language codes, information flows and control flow display and consultation replies according to the user request;
c2: optionally, the system receives feedback comments of the user on a plurality of contents in the display of the configuration file, the programming language code, the information flow and the control flow and the consultation reply, which are generated by the generated artificial intelligence model in the substep C1;
and C3: optionally, the generated artificial intelligent model correspondingly adjusts the output result according to the user feedback opinion and displays the output result to the user again;
and C4: optionally, repeating sub-steps C2 and C3 until the user is satisfied with or terminates the task including several of the configuration file, the programming language code, the information flow and control flow presentation, the advisory reply;
c5: storing a plurality of configuration files and programming language codes so as to facilitate compiling and running of equipment software;
c6: optionally, compiling and deploying the program language code.
7. A device development system based on a generative artificial intelligence model for implementing a device development method based on a generative artificial intelligence model as claimed in any one of claims 1 to 6, characterized in that: the device development sample library is responsible for storing and managing a plurality of sample materials in the device configuration sample, the interface document sample and the configuration display sample;
model training module: training or fine-tuning the generated artificial intelligent model to enable the generated artificial intelligent model to have equipment development related capability;
generating an artificial intelligence model: the system is responsible for analyzing and processing a plurality of tasks including configuration files or program code generation, configuration files or program code modification, configuration display and information consultation, which are proposed by a user, and giving required replies;
and a system integration module: the configuration file generated by integrating and deploying the generated artificial intelligence model is responsible for compiling and deploying the programming language code;
and a system input module: the method is in charge of receiving a plurality of inputs of a user in a mode of containing words, sounds, pictures and videos, and converting the inputs into a format which can be processed by the generated artificial intelligence model;
and a system output module: and the content generated by the generated artificial intelligent model is displayed to the user in a form comprising a plurality of text output, picture output, audio output and video output.
8. The device development system based on a generative artificial intelligence model of claim 7, wherein: sample formats in the equipment development sample library comprise a plurality of characters, program language codes, charts, pictures, models, drawings, sounds and videos.
9. The device development system based on a generative artificial intelligence model of claim 7, wherein: the model training module may comprise any one or more training algorithms for training the generative artificial intelligence model.
10. The device development system based on a generative artificial intelligence model of claim 7, wherein: the generated artificial intelligence model can be obtained by adopting the existing pre-training model and utilizing the device development sample library to finely tune, or can be obtained by utilizing sample data comprising the device development sample library to train.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310482739.XA CN116450100A (en) | 2023-05-03 | 2023-05-03 | Equipment development method and system based on generation type artificial intelligent model |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310482739.XA CN116450100A (en) | 2023-05-03 | 2023-05-03 | Equipment development method and system based on generation type artificial intelligent model |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN116450100A true CN116450100A (en) | 2023-07-18 |
Family
ID=87132046
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310482739.XA Pending CN116450100A (en) | 2023-05-03 | 2023-05-03 | Equipment development method and system based on generation type artificial intelligent model |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116450100A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116756047A (en) * | 2023-08-16 | 2023-09-15 | 江西五十铃汽车有限公司 | Software development method and system of vehicle controller based on GPT |
| CN117131181A (en) * | 2023-10-24 | 2023-11-28 | 国家电网有限公司 | Construction method of heterogeneous knowledge question-answer model, information extraction method and system |
-
2023
- 2023-05-03 CN CN202310482739.XA patent/CN116450100A/en active Pending
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116756047A (en) * | 2023-08-16 | 2023-09-15 | 江西五十铃汽车有限公司 | Software development method and system of vehicle controller based on GPT |
| CN116756047B (en) * | 2023-08-16 | 2023-12-29 | 江西五十铃汽车有限公司 | Software development method and system of vehicle controller based on GPT |
| CN117131181A (en) * | 2023-10-24 | 2023-11-28 | 国家电网有限公司 | Construction method of heterogeneous knowledge question-answer model, information extraction method and system |
| CN117131181B (en) * | 2023-10-24 | 2024-04-05 | 国家电网有限公司 | Construction method of heterogeneous knowledge question-answer model, information extraction method and system |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN116450100A (en) | Equipment development method and system based on generation type artificial intelligent model | |
| US11900256B2 (en) | Deep learning system | |
| Sagerer et al. | Semantic networks for understanding scenes | |
| CN111325664B (en) | Style migration method and device, storage medium and electronic equipment | |
| WO2023241097A1 (en) | Semantic instance reconstruction method and apparatus, device, and medium | |
| US20210383241A1 (en) | Training neural networks with limited data using invertible augmentation operators | |
| CN117152363A (en) | Three-dimensional content generation method, device and equipment based on pre-training language model | |
| CN114187589A (en) | Target detection method, device, equipment and storage medium | |
| CN109445778A (en) | A kind of method and apparatus that the interface auxiliary based on SVG file generates | |
| US20230289509A1 (en) | Parallel mask rule checking on evolving mask shapes in optical proximity correction flows | |
| CN111125550A (en) | Interest point classification method, device, equipment and storage medium | |
| CN113822144A (en) | Target detection method and device, computer equipment and storage medium | |
| CN116612223B (en) | Digital twin simulation space generation method, device, computer equipment and medium | |
| CN110688116A (en) | Image file analysis method, device, equipment and readable medium | |
| CN116012483A (en) | Image rendering method and device, storage medium and electronic equipment | |
| CN117726732A (en) | Reducing false positive ray traversal in bounding volume hierarchies | |
| CN115982363A (en) | Small sample relation classification method, system, medium and electronic device based on prompt learning | |
| CN117726496A (en) | Reducing false positive ray traversal using ray clipping | |
| CN112801201B (en) | Deep learning visual inertial navigation combined navigation design method based on standardization | |
| Peng et al. | Virtual reality in smart city | |
| CN113239994A (en) | Power grid defect detection method and device based on YOLOv4-tiny algorithm, storage medium and electronic equipment | |
| CN116795972B (en) | Model training method and device, storage medium and electronic equipment | |
| RU194188U1 (en) | Program and information complex for creating cartographic terrain models | |
| CN113656873B (en) | Indoor effect graph generation method and device, storage medium and electronic equipment | |
| Roman et al. | Digital image processing-an object-oriented approach |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| TA01 | Transfer of patent application right |
Effective date of registration: 20240508 Address after: 100094, Room 1001, 2nd Floor, 6th Floor, East District, No. 102 Youyi Road, Haidian District, Beijing Applicant after: Xu Yan Country or region after: China Address before: Room 220, 2nd Floor, Building B, No. 22 Xinxin Road, Haidian District, Beijing, 100085 Applicant before: Beijing yizhilian Technology Co.,Ltd. Country or region before: China |