150results about How to "Reduced ability to recognize" patented technology

The invention discloses a photovoltaic cleaning robot based on automatic path planning and an operation method thereof, relates to the technical field of robots, and solves the problem of cleaning blind spots in an existing cleaning robot and the endless loop risk of a path planning algorithm. According to the invention, the surface information of the photovoltaic panel is accurately captured through the multispectral acquisition module, a high-resolution and high-definition comprehensive image is generated through the data fusion processing module, then dirt is accurately identified on different scales by using a multi-scale pyramid structure and an adaptive feature fusion mechanism, and an attention mechanism and environmental perception data are combined, so that the real-time performance of the system is improved. The recognition threshold is dynamically adjusted, and misrecognition and missing detection are reduced; and finally, enabling the robot to autonomously learn in a complex environment and optimize an obstacle avoidance strategy through a dynamic obstacle avoidance algorithm based on deep reinforcement learning. The intelligent level and the cleaning effect of an existing photovoltaic cleaning robot are greatly improved.

The invention relates to the technical field of data analysis, and discloses an intelligent decision and response method and system based on data analysis, and the method comprises the steps: collecting and preprocessing multi-scene monitoring video data based on a security camera, and obtaining preprocessed image data; performing multi-scale feature extraction and L2 regularization feature selection to obtain a target multi-scale feature set; performing scene classification and object recognition to obtain a scene classification result and an object recognition result; performing three-dimensional scene reconstruction and scene graph sequence modeling to obtain a three-dimensional scene model and a scene change prediction result; performing multi-dimensional scene error analysis and reinforcement learning to obtain a plurality of candidate monitoring action strategies; according to the method, adaptive parameter calibration of space-time fusion is carried out, and target monitoring action parameters of the security camera are generated, so that misrecognition and false detection in a complex environment are effectively reduced, and the anti-interference capability of the system is improved.

The embodiment of the invention provides a mining area vehicle path planning method and device based on a digital twin map, and belongs to the technical field of vehicle path planning, and the method comprises the steps: obtaining map data and physical scene data of a mining area; creating a dynamic twin map of the mining area based on the map data and the physical scene data; obstacle information is identified and marked based on the dynamic twin map, and obstacles comprise natural obstacles and artificial obstacles; and importing related data of the vehicle into the dynamic twin map, and carrying out obstacle avoidance path planning on the vehicle based on the marked obstacle information to obtain an optimal obstacle avoidance path. According to the scheme, the running efficiency of mining area vehicles can be improved, and collision and obstacle detouring time is shortened.

The invention discloses an entity relationship extraction method fusing semantic information and syntactic information, and belongs to the technical field of language processing. According to the method, the information in the sentences is learned, so that the fusion of semantic and syntactic information is realized, and the accuracy of relationship classification between entities is improved. According to the method, an external syntax device and word vector information are combined, the influence of noise in a text is reduced by using a graph convolutional neural network of a self-attention pooling layer and a residual shrinkage network, and long-distance dependence between entities is improved. Meanwhile, loss is calculated in an interpolation prediction mode, and interaction between the BERT and the graph neural network is perfected. Experiments show that semantic vectors and syntactic structure information are fully fused, so that the relation extraction effect is remarkably improved compared with other reference models, and the problems that the recognition rate of the long-distance relation between entities in sentences is low and noise exists are solved.

The invention discloses an unmanned warehouse shelf-oriented automatic cargo checking method and system, and the method is characterized in that the method comprises the following steps: employing a target detection algorithm to extract a cargo boundary frame from captured continuous cargo images; for each cargo bounding box, calculating shape features, color features, texture features, size features and local features of the cargo in the cargo bounding box, and splicing feature vectors of the cargo according to the features; pre-defining cargo templates of different types of cargos, matching the cargo templates with the feature vectors of the cargos so as to obtain a category label of each cargo bounding box, and iteratively optimizing the target detection algorithm through a loss function; and tracking the goods extracted by the target detection algorithm by using a Kalman filtering target tracking algorithm, and obtaining the position and quantity information of the goods from the continuous goods images.

The invention relates to the technical field of intelligent power grid control, and discloses an intelligent power grid control system and method based on gesture interaction, and the system comprises a multi-modal data collection and preprocessing module which is used for collecting gesture data of a user, including images, radar and inertial measurement unit data, the gesture data acquisition module is used for acquiring gesture data of a user, filtering, denoising and normalizing the acquired gesture data of the user, and transmitting the processed data to the multi-modal data synchronization and alignment module; and the multi-modal data synchronization and alignment module refers to the output data of the multi-modal data acquisition and preprocessing module, and performs time and space alignment on the multi-modal data through a timestamp synchronization and space calibration technology. According to the method, features of the image, the radar and the inertial measurement unit are subjected to weighted fusion to realize efficient integration of multi-modal information, so that the gesture recognition precision of the system in a complex and variable environment is remarkably improved, and the robustness and the recognition stability of the system are enhanced.

The invention belongs to the technical field of voiceprint recognition, and discloses an intelligent voiceprint recognition method and system based on features. Comprising the following steps: collecting a plurality of voiceprint data samples, and preprocessing each voiceprint data sample to obtain a comprehensive voice sample; constructing a voice directed graph model based on the comprehensive voice samples, and recognizing a core voice sample set from the voice directed graph model; performing feature extraction on the core voice sample set to obtain # imgabs0 # comprehensive feature vectors; obtaining a sample selection strategy by using a clustering algorithm based on the comprehensive feature vector and the core voice sample set; based on the comprehensive feature vector, a voiceprint recognition model is adaptively trained by using a sample selection strategy, and after training is completed, the trained voiceprint recognition model is applied to voiceprint recognition, so that the advantages of deep learning in the voiceprint recognition field are fully played, the voiceprint recognition speed and efficiency are improved, and the voiceprint recognition efficiency is improved. And the misrecognition rate and the missed recognition rate of the system are also reduced.

The invention provides a stock bin grating blockage detection and automatic crushing motion control system based on machine vision, and relates to the field of image data processing, and the system comprises a blockage sensing module which is used for obtaining object sensing data of a stock bin grating based on infrared rays, and judging whether the stock bin grating is blocked or not; the image acquisition module is used for acquiring an image of the storage bin grid according to the object sensing data of the storage bin grid when the blockage sensing module judges that the storage bin grid is blocked; the blockage recognition module is used for judging whether blockage occurs on the stock bin grating according to the image of the stock bin grating and the object sensing data, and determining blockage information of the stock bin grating according to the image of the stock bin grating and the object sensing data when judging that blockage occurs on the stock bin grating; and the ore crushing module is used for generating a crushing scheme according to the blockage information of the stock bin grating and controlling an ore crushing mechanical arm to crush the ore according to the crushing scheme, and the device has the advantage that automatic detection and crushing of blockage of the stock bin grating are achieved.

The invention relates to the technical field of identity verification, in particular to an identity verification method and system based on multi-modal data fusion, and the method comprises the following steps: based on user login information, comparing the login information with pre-stored information, recognizing a user identity, and extracting pre-stored face data and iris data of a user to obtain a user identity information extraction result. According to the method, the accuracy of identity verification is improved by combining face and iris recognition technologies, optimizing image quality and comparing pre-stored data, the possibility of wrong recognition and identity fraud can be remarkably reduced, and flexible response can be performed according to different security requirements by dynamically adjusting a matching degree verification threshold value; according to the method, the adaptability and the protection degree of the verification process are enhanced, the false identification rate and the false rejection rate can be effectively reduced, meanwhile, a stricter security verification process is provided, higher-level personal identity protection is realized, identity embezzlement and other related fraudulent behaviors are effectively prevented, and the trust of consumers to digital services is enhanced.

The invention discloses an RFID antenna power control method and system considering partition identification, electronic equipment and a storage medium. The control method comprises the steps that S1, an RFID antenna transmits an electromagnetic signal to an RFID tag according to a preset transmitting power level; s2, the RFID tag carried by each electric energy meter responds to the electromagnetic signal and receives a reference message fed back by the corresponding electric energy meter; s3, the RFID reader-writer receives and records the RFID data transmission rate, the response signal strength, the corresponding electric energy meter reference information and the transmitting power level of each RFID tag under different transmitting power levels; s4, calculating a characteristic value according to the received data, and calculating a confidence coefficient according to the fed-back quantity of the electric energy meters; and S5, determining the optimal transmitting power level of the RFID antenna by combining the feature values and the confidence coefficients under different transmitting power levels. The problems that an existing RFID system is difficult to achieve accurate fine-grained identification in a complex storage environment, and the complexity of data management is increased due to the signal reflection and interference problems are solved.

The invention belongs to the field of data processing, and particularly relates to an intelligent service analysis optimization method and system based on a multi-dimensional data warehouse, and the method comprises the steps: obtaining the data attribute information of a to-be-recognized object for the to-be-processed object loaded in the data warehouse; inputting the data attribute information into a dynamic data analysis model, and performing data identity recognition on the to-be-recognized object to obtain target identity information of the to-be-recognized object; based on the target identity information, performing identity tracing on the to-be-identified object by adopting a dynamic tracing model to obtain target tracing information of the to-be-identified object; and based on the target identity information and the target traceability information, triggering a target data service for the to-be-identified object to complete intelligent service analysis of the to-be-processed object. According to the method, the data identity recognition and traceability capability can be improved, powerful support is provided for intelligent service analysis, and the method is dedicated to realizing data-driven decision making.

The invention provides an identification and early warning method and system for dangerous behaviors of inland ships. The method comprises the following steps: acquiring infrared monitoring video image data of an inland river water area to obtain a standard data set; then, performing ship main shaft identification according to the standard data set, performing bow and stern identification according to a ship gray value, and performing ship motion direction judgment; then, according to a method based on sample generation, a large number of ship target tracking virtual samples are generated; then, inputting the generated ship target tracking virtual sample into a DVIS model tracking training framework to realize tracking training, and obtaining a tracking model; and finally, according to the obtained tracking model and input real sample data, analyzing motion track data and a motion mode of the ship, and carrying out identification and early warning on dangerous behaviors of the ship. According to the method, tracking of ship behaviors under the complex ship shore environment condition is achieved, early warning is conducted on the ship dangerous behaviors, and the recognition rate of the ship navigation dangerous behaviors is further increased.

The present invention provides a PEG-modified docetaxel liposome lyophilized powder and a preparation method thereof. The composition of the liposome lyophilized powder according to mass percentage is: 0.01%-10% docetaxel; 10%-45% lipid membrane material; the balance is other excipients; the other excipients are PEG materials, antioxidants and lyophilization protectants, and the amount of PEG material added accounts for 0.01%~10% of the total mass of the lyophilized powder. The present invention can improve the deficiencies of ordinary liposomes and injections and other dosage forms by encapsulating docetaxel through long-circulating liposomes, enhance the solubility and safety of docetaxel, change the distribution of docetaxel in the body, and help to achieve good targeting; polyethylene glycol-modified long-circulating liposomes enhance the stability of docetaxel liposomes in plasma, prolong the retention time in the blood, help to exert the EPR effect, and achieve the purpose of high efficiency and low toxicity.

The invention provides an improved convolutional neural network-based sky-bead image automatic identification method and system, and the method comprises the steps: collecting and preprocessing sky-bead images with different patterns and shapes, carrying out the data augmentation operation of the sky-bead images, and building a training set, a verification set and a test set which are needed by a sky-bead image identification model; constructing a convolutional neural network model based on improved OfficientNetV2, introducing a new activation function and an attention mechanism, loading training and verification sets in batches to carry out iterative training on a verification model, and continuously adjusting network parameters to obtain an optimal sky bead image recognition model; inputting a test sky bead image to the optimal model, and returning a sky bead image automatic identification result; according to the method provided by the invention, the accuracy and reasoning speed of sky bead image recognition can be improved on the premise of reducing calculation parameters, a lightweight model can be built for application, the problems of low accuracy, time and labor consumption and the like of traditional manual recognition are solved, and rapid and accurate recognition of the sky bead image can be realized.

The invention belongs to the field of artificial intelligence, and relates to an identity verification method, which comprises the following steps: acquiring user voice and a face image of a user through intelligent glasses; performing voice enhancement processing on the user voice to obtain enhanced voice, and performing identity verification on the user according to the enhanced voice to obtain a first verification result; inputting the face image into an illumination invariant feature extraction model to obtain an illumination invariant feature of the face image, and performing identity verification on the user according to the illumination invariant feature to obtain a second verification result; extracting facial features of the face image, and performing identity verification on the user according to the facial features to obtain a third verification result; and generating an identity verification result of the user according to the first verification result, the second verification result and the third verification result. The invention further provides an identity verification device, computer equipment and a storage medium. According to the invention, the accuracy of identity verification based on the intelligent glasses is improved.

The invention discloses a vehicle driving electronic information system, which comprises an edge computing node, a central control node, a path planning module, a navigation instruction module and a feedback module, and realizes timely response and dynamic strategy adjustment of sudden abnormal events through a real-time data stream processing technology and a reinforcement learning technology. The response speed and flexibility of the system are improved, the data cleaning algorithm and the multi-source data fusion technology are adopted, the multi-source data are processed through the Kalman filtering algorithm and the Bayesian estimation algorithm, and the accuracy and reliability of the data are improved. Through the deep learning technology and the transfer learning technology, the recognition effect under new vehicle models and special conditions is remarkably improved, and the phenomena of misrecognition and missed recognition are reduced. The automatic and intelligent data cleaning and preprocessing reduces manual intervention, improves the efficiency and accuracy of data processing, updates the global model through the federated learning technology, and adapts to data processing requirements in different scenes.

The invention relates to the technical field of intelligent customer service, in particular to a method for improving reply accuracy of intelligent customer service. The method for improving the reply accuracy of the intelligent customer service comprises the following steps: compressing complex statements; vectorization construction of a standard intention library and vectorization construction of a standard entity library; extracting a customer question intention and identifying a key entity, and identifying the intention and the key entity; matching the question intention of the customer with the task; the customer questions the alignment between the key entity and the standard entity library, so that the accuracy of subsequent service parameter matching is improved; automatic assembly and service search of services, automatic assembly of entities, tasks and background services, and accurate query and analysis of intentions are realized; and feedback and display of a result: intelligent and standardized reply is completed based on a large model and a preset template. According to the method, key intentions can be extracted from the long sentences, redundant parts are removed, and the system is enabled to focus on processing core information.

The invention discloses an object tracking method, a mobile intelligent device, a medium and a product, and relates to the mobile intelligent device. The method comprises the following steps: when a user object walks in a target area, receiving an area sound signal collected by the acoustic sensor; extracting a target foot sound signal of the user object from the area sound signal; determining relative position information and walking speed of the user object according to the target footstep sound signal; and controlling the moving direction and the moving speed of the mobile device to be matched with the user object according to the relative position information and the walking speed so as to track the display device. The footstep sound signals of the user are accurately captured and tracked through the acoustic sensor, real-time monitoring of user actions and automatic following of the mobile device are achieved, and interactivity and user experience of the device anytime and anywhere are enhanced.

The invention belongs to the crossing field of artificial intelligence and barrier-free assistance technology, and particularly discloses a sign language translation method and device based on a large language model.The method is applied to a terminal and comprises the steps that key points are extracted from sign language video frames; based on the coordinates, types and global indexes of the key points in the sign language video frames, obtaining fusion features through feature fusion; based on the fusion features corresponding to the sign language video frames in the frame window, time sequence modeling and classification are carried out, and a word prediction result of the frame window is obtained; a sliding window voting and deduplication mode is adopted, the candidate word sequence is preprocessed, a preprocessed word sequence is obtained, and the candidate word sequence is formed by connecting word prediction results of all frame windows in series according to a time sequence; based on the preprocessed word sequence, word error correction and statement generation are carried out by calling a large language model, and a sign language translation result is generated. According to the invention, sign language video translation can be realized for the terminal, and the accuracy and fluency of the translation result can be guaranteed.

The invention discloses a personal voice intelligent terminal for a fund transaction room. The personal voice intelligent terminal comprises a voice data acquisition and classification module, a data cleaning and structuring module, a task concurrency feature extraction module, a concurrency situation prediction module, a situation classification module, a normal situation processing module and a high concurrency situation processing module. And the voice data acquisition and classification module is used for acquiring voice data of traders in real time through terminal equipment in a financial transaction scene. According to the method, the voice data is collected and classified in real time, the feature engineering and the deep learning model are combined, the voice stream processing mode is dynamically adjusted, the high task concurrency situation is effectively dealt with, and misrecognition and resource conflicts are reduced. Voice streams are segmented and fragments are identified independently under the conditions of resource saving and high concurrency in a normal situation, so that major identification errors are prevented, fund loss and strategy failure risks are reduced, the market order stability is guaranteed, and the efficiency and security of a transaction room are improved.

The invention relates to the technical field of voice recognition, in particular to a vehicle voice recognition method and system, and the method comprises the steps: collecting voice data in a cab through a sound pickup, and recording the voice data as initial voice data; detecting and analyzing the initial voice data by using a VAD algorithm, and identifying voice frames in the initial voice data; analyzing each voice frame to obtain defect voice data; repairing the defective voice data, and recording the defective voice data as corrected voice data; evaluating a restoration result based on the corrected voice data; by analyzing the defective voice data and generating the corrected voice data, the system can optimize the recognition result and reduce the risk of misoperation, and meanwhile, by judging the accuracy of the corrected voice data, whether to directly execute an instruction or return confirmation can be intelligently determined, the user experience and the operation safety are improved, and the user experience is improved. According to the method, improper operation caused by voice recognition errors can be reduced, and particularly in the driving process, safety can be better guaranteed.

The utility model belongs to the technical field of walking equipment, and particularly provides a mobile robot which comprises a machine body and a side face sensing assembly, the machine body can move on a working face, and the side face sensing assembly is arranged on the side portion of the machine body. The side face sensing assembly is used for detecting environment information of the side of the machine body. The side sensing assembly is arranged at the side part of the machine main body, so that the environment information of the side of the machine main body can be directly acquired, and the environment information of the side does not need to be indirectly calculated by analyzing and disassembling images of other sensors at the front or the top; therefore, the identification and following difficulty of the lateral operation boundary of the machine main body can be reduced, the overall calculation amount of the mobile robot in the moving process is effectively reduced, the reaction efficiency of the mobile robot is improved, and the moving reliability is improved.

The application provides an antibacterial neural electrode based on a high-surface-potential zirconium oxide layer and a preparation method thereof. The electrode comprises, from bottom to top, a bottom zirconium oxide film layer, a first insulating layer, a metal layer, a second insulating layer, a top zirconium oxide film layer and an electrode modification layer. A distal end of the metal layer is partially exposed to form an electrode point, and the electrode modification layer covers the electrode point. The bottom zirconium oxide film layer and the top zirconium oxide film layer are subjected to high-surface-potential treatment and exhibit a positive surface potential in a physiological environment. The positive surface potential can destroy the cell membranes of bacteria with negative charges through electrostatic action after the neural electrode is implanted in vivo, and inhibit the formation of microglial cells and astrocytes. The damage to the membranes of mammalian cells (such as neurons) with neutral electricity is extremely small. The application realizes active antibacterial function through the high-surface-potential zirconium oxide film layer, ensures long-term effective contact between the electrode and neurons, and thus realizes high-quality, long-term stable electrophysiological recording and stimulation function of the neural electrode.

The invention relates to the field of image processing, and provides a method for quickly positioning abnormal visual information in a spliced image, which comprises the following steps of: marking all spliced areas in the spliced image, sequentially segmenting each spliced area into a plurality of sub-areas, and positioning abnormal image blocks according to pixel distribution in the sub-areas. And highlighting all abnormal image blocks in the spliced image. According to the method, abnormal visual information in the spliced image can be effectively identified and positioned, the requirement of manual participation in the identification and screening process is reduced, pixel distribution in the sub-regions is analyzed in parallel, a larger amount of image data can be processed in a shorter time, the overall visual quality of the spliced image is remarkably improved, and the image quality is improved. And the method can be widely applied to various images needing to be spliced, and has relatively high universality.

The invention relates to the field of data analysis and digital finance, and discloses a text matching method and device based on a word segmentation device and a dictionary tree, equipment and a medium, and the method comprises the steps: determining an input word segmentation fragment based on an initial input text and the word segmentation device; taking the input word segmentation segment as a word segmentation index, and matching the word segmentation index with each text in a preset dictionary tree; and determining a preset text matched with the word segmentation index as a target text matched with the initial input text. By means of the mode, the preset word segmentation device and the dictionary tree are combined for use, vocabularies and phrases in the text can be recognized and matched more accurately, and therefore the possibility of misrecognition and mismatching is reduced. Through the preset dictionary tree, searching and matching are quickly performed in a large number of preset texts, so that the efficiency of the whole word segmentation and matching process is improved, and the accuracy of performing text matching according to the keyword input by the user in the financial field is improved.

This application provides a structured processing method, apparatus, electronic device, and medium for software engineering elements, relating to the field of software development technology. The method includes: generating an engineering element table for multiple types of software engineering objects for a preset software engineering project; each engineering element in the table records an element identifier and basic attribute information of the engineering element in the software engineering object; constructing an element reference chain based on the relationships between multiple engineering elements and element identifiers; the element reference chain consisting of at least a source element identifier, a target element identifier, and a relationship type; responding to a change event for a first engineering element, determining the affected engineering element based on the element identifier and element reference chain of the first engineering element; and verifying the affected engineering element based on a verification list generated for the affected engineering element. This application can determine affected engineering elements based on unified engineering elements and element reference chains, improving the efficiency of software development and version release.

The invention relates to a gesture recognition method and device, computer equipment and a storage medium. The method comprises the following steps: acquiring gesture space information of a current gesture based on a received gesture recognition instruction; the gesture space information comprises three-dimensional coordinates of fingers and a wrist; obtaining a first serialization feature of the current gesture based on the three-dimensional coordinates of the fingers and the wrist; sequentially calculating gesture matching degrees of the first serialized features and second serialized features of a plurality of preset gestures; and determining a gesture recognition result of the current gesture based on the gesture matching degree. By adopting the method, the matching of the current gesture and the preset gesture can be realized according to the processing of the characteristics of fingers, wrists and the like, the gesture recognition result can be obtained without a neural network or machine learning training recognition, and the effect of improving the gesture recognition speed and accuracy is achieved.

The application discloses a human-computer voice interaction method and device, a storage medium and an electronic device, relates to the technical field of smart homes, and comprises the following steps: automatically performing voice recognition on voice interaction information of a target object, obtaining a first intermediate recognition text and a second intermediate recognition text, a pause duration threshold set in a first voice activity detection strategy for detecting a moment when the target object stops speaking is lower than a pause duration threshold set in a second voice activity detection strategy for detecting the moment when the target object stops speaking; comparing the first intermediate recognition text and the second intermediate recognition text until the second intermediate recognition text is consistent with the first intermediate recognition text, obtaining a target recognition result of the voice interaction information based on a voice recognition end identifier of the monitored voice interaction information; and performing natural language processing on the target recognition result to obtain a first interaction intention, thereby solving the technical problem that automatic voice recognition is not accurate enough in the human-computer voice interaction process.

The invention provides an automatic cone searching method based on magnetic field sensing, which comprises the following steps of: acquiring real-time magnetic induction data output by a magnetic resistance sensing unit arranged on a cone searching device based on a target magnet, and a magnet position corresponding to the target magnet based on the real-time magnetic induction data; generating a moving direction for controlling the cone searching device to move towards the target magnet based on the magnet position, and triggering a cone barrel recognition unit arranged on the cone searching device to perform feature collection corresponding to the moving direction when the cone searching device moves based on the moving direction to obtain real-time feature data; and cone barrel feature data corresponding to the cone barrel are called, the target magnet is determined as a magnetic element arranged in the cone barrel under the condition that the real-time feature data are matched with the cone barrel feature data, and a mechanical arm arranged on the cone searching device is controlled to execute collection operation on the cone barrel corresponding to the target magnet. According to the invention, the efficiency and precision of automatic operation are at least improved.

The invention discloses a control method and a control system of a smart home system, relates to the technical field of smart homes, and aims to solve the technical problem that a control system cannot recognize a corresponding instruction or mistakenly recognizes the corresponding instruction due to current traditional voice instruction control, and further causes unnecessary troubles. Comprising the following steps: S1, defining a multi-dimensional and refined film watching behavior mode; comprising body postures, facial expressions, eyeball movement, sound features, handheld device operation and the like, specific recognition standards and threshold values are set for each type of features, and a film watching behavior pattern library is formed; s2, behavior pattern acquisition; by integrating various high-precision sensors such as a high-definition camera, an infrared sensor, a pressure sensor and a microphone, multi-dimensional behavior data of a user in the film watching process are collected in real time and preprocessed to improve the data quality; and S3, behavior pattern recognition. The method has the advantage of improving the convenience and accuracy of intelligent home equipment control.