8 results about "Dual code" patented technology

The invention belongs to the technical field of new-generation information, and particularly relates to a power plant metal supervision entity relationship extraction method based on dual coding, which comprises the following steps: multi-modal data preprocessing: collecting and cleaning data, and carrying out data labeling; constructing a dual coding joint learning model: designing a network layer architecture, and training the dual coding joint learning model; constructing and querying a dynamic knowledge graph: extracting a model to generate a triple, and performing time sequence evolution analysis, causal reasoning interface and dynamic updating; and incremental knowledge updating and dynamic model optimization: an incremental learning and feedback module forms a bidirectional closed loop between the knowledge graph and the relationship extraction model, and continuous evolution of the system is ensured through dynamic knowledge updating and adaptive model optimization. According to the method, through bidirectional feature modeling of dual paths, collaborative representation optimization between entities and relationships is realized while context information is captured, so that the requirements of complicated data types and diversified semantic associations in an engineering scene are met.

The present application belongs to the technical field of information security and coding theory, and provides a minimal linear code construction method, a secret sharing method and a secret sharing system, comprising: first defining a trace function in a finite field, constructing a two-to-one mapping, a vector Boolean function with a finite discrete value of Walsh spectrum, and then defining a code word set based on the function, embedding the output of the trace function into a linear code generation expression to generate a minimal linear code with a weight set and a support without strict inclusion; then using the dual code to establish a one-to-one correspondence between the participants and the dual code coordinates, the minimal code word support and the minimum access structure, embed the secret generation share distribution, and the participants meeting the conditions restore the secret through linear combination. The system contains four functional modules, which are suitable for distributed storage, blockchain key management and other scenes, and realize safe, efficient and expandable secret sharing. The present application can improve the security, efficiency and flexibility of secret sharing, and is suitable for multiple scenes and easy to expand.

The invention relates to a patch overfitting evaluation method combined with dual code representation, and belongs to the field of internet security services. Aiming at an automatic patch generation condition, firstly, a defect code snippet, a correct code snippet and CFG and AST of a patch are constructed, and a Word2Vec model is utilized to respectively generate embedded vectors containing structural information and semantic information; secondly, a Transform model is utilized to encode an AST embedded vector, a GAT model is utilized to encode a CFG embedded vector, and a feature vector is obtained through splicing; then calculating feature vector distances between the patches and the defect code snippets and between the patches and the correct code snippets, and training to generate a dichotomy model for calculating patch overfitting confidence; and finally, weighting the confidence coefficient according to the code change degree, and comparing the weighted confidence coefficient with a threshold value to judge whether overfitting exists. Aiming at the defects in the existing method, the semantic similarity of correct code snippets is high, and misjudgment is easily caused when the code change degree is large, the code CFG features are increased, the code change degree is calculated, and the patch overfitting evaluation accuracy is improved.

To provide a method for increasing the efficiency of residual coding.SOLUTION: An image decoding method includes the steps of obtaining a dependent quantization enabled flag, obtaining a TSRC available flag on the basis of the dependent quantization available flag, determining residual coding syntax for the current block on the basis of the TSRC available flag, obtaining residual information on the determined residual coding syntax for the current block, deriving a residual sample for the current block on the basis of the residual information, and generating a reconstructed picture on the basis of the residual sample, and the dependent quantization available flag is a flag for whether dependent quantization is available, the TSRC available flag is a flag for whether the TSRC is available, and the TSRC available flag is obtained on the basis of the dependent quantization available flag whose value is 0.SELECTED DRAWING: Figure 10

This invention relates to the field of battery safety monitoring technology, specifically to a method and system for tracing batteries after combustion based on heterogeneous dual codes, solving the technical problem that existing technologies struggle with identification and tracing under abnormal conditions. The method includes: acquiring a microscopic depth image of the battery tab surface, performing image depth analysis based on the microscopic depth image to determine the character sequence corresponding to the plaintext character distribution area on the battery tab surface and the set of etching point coordinates corresponding to the QR code distribution area; performing database traversal matching based on the character sequence to obtain at least one candidate factory identification code and a QR code bitmap corresponding to each candidate factory identification code; and performing structural comparison analysis based on the etching point coordinate set and the QR code bitmaps corresponding to each candidate factory identification code to trace back to the target factory identification code from the at least one candidate factory identification code.

The invention relates to the technical field of sample storage, particularly provides a cryopreservation box code scanning device and a control method of a sample storage library, and aims to solve the problem that an existing single code scanning structure is low in code scanning efficiency. In order to achieve the purpose, the cryopreservation box code scanning device comprises a base, a box carrying table and a code scanning mechanism installed on the base. The code scanning mechanism comprises a code scanning box, a first code scanning camera and a second code scanning camera, a first code scanning station and a second code scanning station are arranged on the code scanning box, and the first code scanning camera and the second code scanning camera are arranged on the first code scanning station and the second code scanning station correspondingly and can conduct code scanning operation on the cryopreservation boxes; the box carrying table is arranged above the code scanning box, a first box carrying station and a second box carrying station which are used for carrying the cryopreservation boxes are arranged on the box carrying table, the first box carrying station corresponds to the first code scanning opening, and the second box carrying station corresponds to the second code scanning opening. By adopting the double code scanning stations, code scanning operation can be carried out on the cryopreservation boxes at the same time, and the code scanning efficiency is improved.

The invention discloses a pipeline parallel fault tolerance method based on erasure codes and a related device. Belongs to the technical field of deep learning. The method comprises the following steps: firstly, initializing a redundant block by adopting a dual coding strategy before training is started, and then updating parameters of the redundant block before all-reduce operation to ensure that the redundant block and model parameters are synchronously updated, so that the redundant block is recovered after a fault occurs; and finally, when the node fails, executing a recovery process based on the data block and the redundant block of the existing node, and recovering complete model parameters through the existing data block and the redundant block. Therefore, an assembly line parallel training fault-tolerant strategy with low redundant storage overhead and high recovery efficiency is realized.

A dual-code lock comprises a lock housing (15), an actuating element (11, 11') for adjusting a lock bolt, at least one code wheel (3) rotatably mounted in the lock housing (15) about a pivot axis (A), a first and second setting bushing (4, 5) arranged coaxially to the code wheel (3) on opposite end faces, each with a coding notch (43, 53) in a cylindrical surface of the setting bushing (4, 5), two rockers (6, 7) partially encompassing the at least one code wheel (3), each with a pivot axis (62, 72) pivotally attached to the lock housing (15), a locking axis (61, 71) extending parallel to the pivot axis (62, 72), and a crossbar (63, 73) resting on the cylindrical surface of the setting bushings (4, 5), wherein the rockers (6, 7) are positioned between a Actuating element (11, 11') releasing position when the first or second code wheel (3) is entered with at least one code wheel.second locking secret and a locking position that locks the actuating element (11, 11') is pivotable, wherein all actuating bushings (5) are detachably and rotationally fixed mounted on the respective code wheel (3).