28 results about "Class Code" patented technology

According to the PCIe retimer, the PCIe system and the electronic equipment, a decoding and descrambling module carries out decoding and descrambling processing on output code streams of upstream equipment and outputs the code streams backwards, and when a low-delay triggering condition is met, a control module modifies second-class code streams in the decoding and descrambling code streams into first-class code streams and outputs the first-class code streams backwards; the de-skew module performs de-skew processing on the received code stream and outputs the code stream backwards; the coding and scrambling module carries out coding and scrambling processing on the received code stream and backwards outputs the code stream subjected to coding and scrambling processing; and the control module controls the low-delay access to serve as the current working access after determining that the low-delay access is aligned with the main access. According to the embodiment of the invention, the state machine of the downstream device does not perform state hopping due to receiving the preset number of second type code streams, so that the state jump time of the upstream device and the downstream device can be dragged, the purpose of dragging the state machine is achieved, and the PCIe retimer has more time to perform path switching without error codes.

The invention provides a robot control method and system based on a big language model, a double big language model evaluation mechanism is adopted, a big language model of a code generation and modification module uses Claide 3.5 sonnet, control codes are generated by the Claide 3.5 sonnet, evaluation is provided by adopting a gpt-4o evaluation model of OpenAI to optimize suggestions, and the big language model of the code generation and modification module is used for evaluating the control codes. The risk that problems are difficult to find due to homogenization of the same module is reduced, and the generation quality, safety and supportability are improved; in addition to a double large language model evaluation mechanism, due to the fact that safety prompt words are added into prompt words of a generation model and a generation template example is provided, operation of the generation model in a specific framework can be limited, the possibility of error occurrence is reduced, and specification and safety of a generation effect are further provided; in addition, the control class verification module constructs a simple simulation environment after the control class codes are generated, so that grammar errors and logic vulnerabilities of the generated codes can be screened in advance, and potential safety hazards are further reduced.

The invention provides a code peculiar smell detection method and system based on a large language model, and relates to the technical field of code peculiar smell detection.The method comprises the steps that context code information of a to-be-detected code is extracted, and the context code information comprises a method code and a class code; respectively extracting key information of the method codes and the class codes, and constructing a method information set and a class information set; constructing a large language model; on the basis of a thinking chain and a thinking tree, constructing cue words of the large language model; constructing a hierarchy-based code replacement strategy; dynamically adjusting the context code information by using the method information set and the class information set according to the code replacement strategy; and inputting the cue word and the dynamically adjusted context code information into the large language model for code odor detection, and outputting a code odor detection result of the to-be-detected code.

The invention provides a camouflage object semantic segmentation method and device based on an adaptive candidate strategy, and a medium, and the technical scheme of the invention obtains the category code of a to-be-segmented image through a coarse feature extractor and a classifier, achieves the transmission of semantic category information, and enhances the semantic perception capability of a subsequent segmentation task. Initial prediction of a camouflage object is carried out through a target detector and an adaptive candidate strategy, an optimal target attention box is generated, and robust space guidance is provided for a core segmentation task; according to the method, category coding and a target attention box are combined, multi-source information is fused and reconstructed through a multi-guide feature fusion module and a multi-task perception decoder, and an accurate segmentation result is output.

This invention discloses an automated sports equipment management system and method based on machine learning, comprising the following steps: acquiring borrowing and returning trigger data, class code image acquisition, ball return area image acquisition, ultrasonic distance data, and equipment quantity data; inputting the class code image into an improved YOLOv2 ball borrowing and returning status recognition model deployed in a K210 vision module to obtain class recognition results; controlling the servo motor to release the ball based on the class recognition results and the number of remaining balls; opening the ball return cover and acquiring ball return detection data when returning the ball; generating a ball return confirmation result by combining ultrasonic distance changes and ball target recognition results; updating the number of borrowed balls in the class and the number of remaining balls in the equipment, and generating display, voice, and recording data. This invention achieves automated borrowing and returning management of ball sports equipment based on machine learning visual recognition and sensor linkage control technology.

Systems, methods, and other embodiments associated with a centralized product catalog with standardization from independent data sources are described. In one embodiment, a computing system is configured to access and download product information for a plurality of products that is maintained in a non-standard format. Metadata is identified and extracted. Feature vectors may be generated from the metadata and used to predict and assign class codes to products. The product attributes are extracted from each product, which include a non-standard product description. The non-standard product description is converted into a standardized text description. Standardized records are generated for each product record and a centralized database is generated with the standardized records.

The disclosure provides a camouflage object semantic segmentation method and device based on an adaptive candidate strategy, and a medium. The technical scheme of the disclosure obtains the class code of a to-be-segmented image through a coarse feature extractor and a classifier, realizes the transmission of semantic class information, and enhances the semantic perception ability of a subsequent segmentation task. An initial prediction of a camouflage object is performed through a target detector and an adaptive candidate strategy, and an optimal target attention box is generated to provide robust spatial guidance for a core segmentation task. The class code and the target attention box are combined, multi-source information is fused and reconstructed through a multi-guided feature fusion module and a multi-task perception decoder, and an accurate segmentation result is output.

The application discloses a software design and development method and a terminal, a detailed design document corresponding to an architecture design document is generated according to software requirements, a corresponding core class code is generated according to a core class diagram of the detailed design document, and an engineering code is obtained by directly developing based on the core class code by a developer; if the engineering code does not match the core class in the detailed design document, the core class diagram and the corresponding design document are re-generated, thereby greatly improving the consistency of software design and software development; if the engineering code matches the core class of the detailed design document, the engineering code is submitted to a code warehouse, the submitted document is ensured to be correct and reliable, and the design document is ensured to be well landed in the development stage.

The invention relates to a neural symbol multi-hop reasoning method and system for a domain knowledge graph. The method comprises the following steps: dividing a domain knowledge graph into multi-order sub-graphs, and performing symbol reasoning on each multi-order sub-graph according to a chained domain ontology to obtain an extended sub-graph; taking each multi-order sub-graph as a training sample, and coding instances in the multi-order sub-graphs to obtain class coding vectors and assertion coding vectors; constructing an assertion set reasoning network; taking assertions in the extended subgraph and the negative example set as a positive example and a negative example to construct a supervision label, and training an assertion set reasoning network by using a training sample and the supervision label; dividing a to-be-inferred domain knowledge graph into to-be-inferred multi-order sub-graphs, performing instance coding, inputting the to-be-inferred domain knowledge graph into the trained assertion set inference network, and obtaining corresponding extension sub-graphs according to the output extension type assertion and the established relation assertion; and combining the extended sub-graphs to output the extended domain knowledge graph. By adopting the method, a high-quality, strong-generalization and high-efficiency chain domain knowledge graph reasoning task can be realized.

The invention discloses a code abstract generation method fusing item-level code context, and belongs to the field of code abstract generation. Comprising the following steps: acquiring a code library to be subjected to abstract generation in a Java language, and acquiring a previous text, a next text and an abstract syntax tree language fragment of each segment of code in the code library; obtaining basic abstracts of various codes in the code library based on the trained model; obtaining context selection strategies of various codes based on the trained context selection strategy identification model; acquiring a context required by the class code from the context of the class code based on the context selection strategy of each segment of class code; and filling a preset prompt template with the basic abstract, the abstract syntax tree language segment, the required context and the metadata of each segment of class code, and inputting the prompt template filled with information into a large language model to obtain code abstracts of all class codes in the code library. According to the method, the code abstract with complete information and higher readability and accuracy can be generated.

The present application discloses a method, apparatus, device, and storage medium for determining a transfer learning model, belonging to the field of computer and Internet technologies. The method includes: determining a plurality of candidate network layers from at least one candidate transfer learning model; processing a training sample set based on the candidate network layers to obtain a sample coding information entropy and class coding information entropies respectively corresponding to a plurality of classes; determining a transfer rate of the candidate network layers according to the sample coding information entropy and the plurality of class coding information entropies; and constructing a transfer learning model for the training sample set based on the candidate network layers respectively corresponding to each transfer rate and according to the candidate network layers whose transfer rates satisfy a first condition. In the present application, a method for evaluating the transfer learning effect before transfer learning is provided, which can evaluate the transfer learning effect of the candidate network layers on the training sample set without transfer learning, so as to quickly and accurately determine the optimal candidate network layer for the training sample set.

The application provides a GIS-based township land space planning base processing method and system, which comprises the following steps: extracting the past land class code from the three-adjustment plot vector to obtain direct conversion vectors and indirect conversion vectors, and updating the direct conversion vectors into first updated three-adjustment plot vectors; updating the indirect conversion vectors into second updated three-adjustment plot vectors and to-be-planned vectors through three types of data; updating the to-be-planned vectors into third updated three-adjustment plot vectors, and then obtaining final three-adjustment data. Compared with the manual processing method, the direct conversion vectors are converted into the first updated three-adjustment plot vectors, which avoids repeated calculation of most stable land blocks, and the coding of the indirect conversion vectors can be automatically assigned, which improves the base processing efficiency. Moreover, the three types of data with dynamic and real-time characteristics are introduced to make up for the tardiness of the three-adjustment data and improve the accuracy.

The application relates to the field of data processing and communication technology, and discloses a standardized implementation method of a dynamic automatic translation device, which comprises the following steps: constructing and analyzing a standard configuration model, generating an attribute dependency matrix, performing loop detection and topological sorting, generating a linear ordered execution sequence, presetting a class code skeleton and a runtime context, extracting a logic code set according to the sequence, applying a code synthesis equation to generate a composite logic source code, calling an instant compiler and an isolated class loader to convert the source code into a dynamic translation instance and perform atomic reference replacement, and feeding device original data into the instance to make the device original data flow in a logical path in a one-way manner in the runtime context to complete translation. Through the conversion of the interpretive execution into the compiled sequential execution, the application eliminates runtime dependency lookup, cooperates with the isolated loading and atomic replacement mechanism, realizes the independent running and smooth hot updating of the device logic, and improves the efficiency and stability of the heterogeneous protocol adaptation.

The invention discloses a Qt code generation method for monitoring source simulation software based on a UML class diagram and a time sequence diagram. The method comprises the following steps: constructing the UML class diagram and the time sequence diagram of the monitoring source simulation software; obtaining an XMI file corresponding to the class diagram and the time sequence diagram; constructing a set based on the XMI file; generating class codes of the Qt; generating an object calling relation of the Qt; generating a signal and slot code; and generating a Qt code file. According to the method, the model of the monitoring source simulation software is specifically applied to the UML class diagram and the time sequence diagram, the process of establishing the model of the monitoring source simulation software is simplified, the design is clear, the structure is indirect, and the efficiency from the model to code generation is better.

Provided in the present application are an installation package generation method and apparatus, code obfuscation methods and apparatuses, and a code obfuscation system. One specific embodiment of the installation package generation method comprises: obfuscating original entity class code among original installation code of an application program to obtain obfuscated entity class code; and generating an installation package of the application program on the basis of the obfuscated entity class code. The method can reduce the size of part of code in the installation package of the application program, and hides details of the original installation code, so that the difficulty of reverse engineering is increased.

The invention discloses a defect positioning method and system based on a large model standard operation process and a storage medium. The method comprises the following steps: step S1, defect understanding; performing byte code instrumentation on the failed test case, recording a method call chain and coverage class information, extracting a test tool class code and document metadata by adopting static analysis, enabling a large language model to integrate an error log, test output and an execution track for analysis, and generating a potential error cause hypothesis; s2, code library navigation; identifying all related methods from the whole code library by using a document enhanced hierarchical code search mechanism; and S3, defect verification. The positioning system and the storage medium are implemented based on the method. The method has the advantages of simple principle, higher defect positioning precision, higher efficiency and the like.

The invention provides a test case generation method and device, electronic equipment, a medium and a program product, relates to application of a large model in code testing, and can be applied to the technical field of artificial intelligence. The method comprises the steps of obtaining at least one target class code, and determining a target method of at least one to-be-generated unit test case from the target class code; for each target method, obtaining a corresponding target method code and historical test information, and constructing model input content; inputting the model input content into a large model to obtain a unit test case; and running the unit test case, obtaining test feedback of this round, and judging whether the unit test case meets a preset generation termination condition or not based on the test feedback of this round.

This application discloses a code generation method and related devices, belonging to the field of code generation technology. The method comprises: obtaining generation description information corresponding to the class code to be generated; inputting the generation description information into a code generation model to obtain the target class code output by the code generation model, wherein the code generation model learns similarities between the class code description information and the abstract syntax tree. In this application, the accuracy of class code generation is improved.

The invention discloses a zero-intrusion-oriented compiling period type conversion generation method. The method comprises the steps that a type conversion annotation in a source code is scanned and analyzed, a generation task is constructed for the type conversion annotation, and the generation task indicates annotated element information and a target type expression; for each generation task, respectively constructing a source type descriptor and a target type descriptor for the annotated element and the target type expression, the source type descriptor and the target type descriptor being based on the same data model; performing consistency verification on the source type descriptor and the target type descriptor in each generation task; and for the generation task which can be subjected to safety conversion according to the consistency verification judgment, a converter class code with a safe type is generated, safety enhancement logic based on a type conversion annotation is embedded in the converter class code, and safe type conversion is realized when the converter class code is called. According to the embodiment of the invention, on the premise of not changing service logic, the converter class code with the safe type can be automatically generated.

Systems, methods, and other embodiments associated with a centralized product catalog with standardization from independent data sources are described. In one embodiment, a computing system is configured to access and download product information for a plurality of products that is maintained in a non-standard format. Metadata is identified and extracted. Feature vectors may be generated from the metadata and used to predict and assign class codes to products. The product attributes are extracted from each product, which include a non-standard product description. The non-standard product description is converted into a standardized text description. Standardized records are generated for each product record and a centralized database is generated with the standardized records.

The application is suitable for the technical field of machine learning, and provides a code representation generation method for class code, which comprises the following steps: performing data reduction on a class-level source code file by adopting a class-level code data reduction strategy to obtain a first sequence and a second sequence after reduction; and performing representation fusion on a first vector representation sequence corresponding to the first sequence and a second vector representation sequence corresponding to the second sequence by adopting a representation fusion strategy based on an attention mechanism to obtain a code representation of the class-level source code file, so that the generation quality and effectiveness of the code representation of the class code are improved, and the model performance is improved.

The invention provides a database entity refreshing method and system and a storage medium. The method comprises the following steps: monitoring table structure change of a database to obtain structure metadata of each data table and a corresponding last change timestamp; each last change timestamp is compared with a corresponding refreshing timestamp of a local cache, and a target data table with a changed structure is judged based on comparison; triggering a dynamic code generation process for the target data table in response to the target data table; dynamically generating an entity class code and a data access layer component code corresponding to the target data table based on the target structure metadata of the target data table and a predefined code generation template; loading the entity class code and the data access layer component code into an application running environment through a dynamic class loading mechanism to replace or register the corresponding class definition; and updating the refresh timestamp information corresponding to the target data table in the local cache after dynamic class loading is completed. The database can be efficiently maintained at low cost.

The invention relates to an application ID code generation method and device, equipment and a storage medium. The method comprises the following steps: receiving a coding instruction of a client; determining a corresponding coding category according to the coding instruction; the coding category comprises a first type of coding and a second type of coding; acquiring corresponding code identification information according to the code category; the code identification information comprises first-class information and second-class information; the first-class information corresponds to the first-class code and comprises a timestamp of current system time, server IP information, a machine code, a current application process number and a random number; the second-class information corresponds to the second-class code and comprises the timestamp, the server IP information and a self-increasing serial number; and generating a corresponding application code according to the code identification information, and returning the application code to the client. The corresponding generation mode is matched according to the service requirement, and the flexibility of code generation is improved.

The application discloses a kind of application reinforcement method and device, storage medium and electronic equipment. Among them, the method comprises: obtaining the code relationship diagram of target application program, wherein the code relationship diagram includes the reference relationship between each class in the bytecode assembly file corresponding to target application program;Based on the class code of each class, generate code retrieval strategy, wherein the code retrieval strategy includes a set of retrieval rules generated using the reference relationship between each class;Based on a set of retrieval rules, determine the code set to be processed from bytecode assembly file;The code in the code set to be processed is reinforced, and the reinforced target application program is obtained. The present application solves the technical problem of low efficiency in the application program reinforcement process in the related art.

The invention discloses a code generation method and related equipment, and belongs to the technical field of code generation. The method comprises the steps of obtaining generation description information corresponding to a to-be-generated class code; the generation description information is input into a code generation model, target class codes output by the code generation model are obtained, and the code generation model learns similar features between description information of the class codes and the abstract syntax tree. According to the method and the device, the class code generation accuracy is improved.

Embodiments of the present application disclose an image processing method, device and equipment and medium, which are applied to the technical field of image processing. The method comprises: obtaining a target image containing a target object, an image feature vector and an object query vector; generating an object feature vector based on the object query vector and the image feature vector; obtaining a class code vector corresponding to each object class in an object class set; determining the similarity between the object feature vector and the class code vector corresponding to each object class, so as to determine the belonging object class of the target object from the object class set. The embodiments of the present application can improve the accuracy of object segmentation in an image.

The invention discloses a software cross-architecture migration workload evaluation method, which relates to the technical field of software, and is characterized by comprising the following steps of: matching an incompatible code segment to be transformed; dividing the incompatible to-be-transformed code segment into a quantum simple transformation type code and a quantum complex transformation type code according to the code segment modification quantum complexity, the associated file quantum entanglement depth, the modification scheme quantum superposition state probability and the quantum tunneling effect influence degree; the quantum basic workload of the class of quantum simplicity problems, the quantum learning efficiency factor of the quantum iteration period, the quantum learning curve curvature coefficient and the quantum iteration period number to which the class of quantum simplicity problems belong are calculated according to the quantum learning efficiency factor and the quantum learning curve curvature coefficient of the class of quantum simplicity problems; constructing a quantum ecological fragmentation coefficient and a quantum time expansion correction coefficient, and carrying out processing calculation to obtain the workload of a quantum simple transformation type code; the basic parameter set is set for the quantum complex transformation type codes, and the effect is that the workloads of different code transformation scenes can be accurately achieved.

The present disclosure provides an annotation processing method, apparatus, device, storage medium, and program product, relating to the field of computer science. The annotation processing method includes: obtaining a code file to be processed, the code file including entity class code, and the entity class code including field information; if it is determined that first annotation information associated with the field information exists in the entity class code, obtaining second annotation information associated with the field information from a database table corresponding to the entity class code; if it is determined that the second annotation information is different from the first annotation information, updating non-target annotation information based on the target annotation information. The present disclosure provides an automated annotation information processing solution, which improves the efficiency of annotation information processing during code development.