45 results about "Pattern recognition system" patented technology

The invention relates to the technical field of computer vision, and discloses a high-speed human motion trail motion mode recognition system. Through multi-modal sensor fusion and an adaptive space-time attention network (ASTAN), the problems that a traditional optical system depends on mark points and IMU accumulative errors exist are effectively solved, and the high-speed movement track reconstruction precision and the environmental adaptability are remarkably improved; self-supervised data enhancement and dynamic model compression technologies are combined, so that the dependence on labeled data is greatly reduced, lightweight edge deployment is realized, and the real-time requirement is met; the system can synchronously output multi-mode feedback instructions, supports multi-scene application such as sports action correction, medical rehabilitation evaluation, security and protection anomaly detection and man-machine interaction control, and solves the core pain points of insufficient precision, real-time performance, robustness and generalization ability in the prior art. And an efficient, reliable and low-cost comprehensive solution is provided for high-speed motion analysis.

The invention relates to a radar working mode recognition system and a semi-supervised training method. The system comprises a supervising module and a self-supervising module which are updated alternately. During training, the supervision module adopts annotation data to drive feature extraction and classifier optimization, and captures high-order feature differences by fusing polynomial loss, so that the model discrimination capability is improved; the self-supervision module innovatively introduces a dynamic time slice division mechanism, adjusts a time slice division proportion in real time according to a verification set index, and forms an elastic training strategy. The two modules share the weight of a feature extraction network, and bidirectional gradient transmission is realized in iteration: feature parameters updated by supervision training are synchronized to the self-supervision module, and the self-supervision module optimizes a generated pseudo-label relation loss feedback feature space by constructing positive / negative samples. According to the collaborative mechanism, labeled data and unlabeled data are aligned in a unified feature space, and the error accumulation problem of a traditional training method is effectively relieved. According to the method, the radar working mode recognition capability in a complex electromagnetic environment can be effectively improved.

The invention relates to the technical field of artificial intelligence, in particular to an anti-money laundering mode recognition system based on AI, which comprises a data acquisition processing module, a data enhancement module, a training optimization module, an early warning evaluation module, a data learning module and a data generation module. A traditional anti-money laundering system usually depends on a static rule set, and the rules are often lagged in facing continuously evolving money laundering means, however, the system provided by the invention has the characteristic of dynamic adaptability by continuously learning new data and updating and optimizing an identification model. The method can dynamically adapt to a new money laundering mode and a complex transaction behavior, and the flexibility enables the system to maintain an efficient recognition capability when facing an unknown threat.

A system and method for a laryngoscopic device which incorporates the electronic and imaging elements within the device with advanced functional iterations that incorporate mathematical pattern recognition / predictive modeling as well as parallel computing and supercomputing for data analysis as well as and integrating it with the patient's record and compiling images from multiple scopes in a cloud-based system for an AI platform to use in diagnosis and disease monitoring.

Methods, devices and systems for training a pattern recognition system are described. In one example, a method for training a sign language translation system includes generating a three-dimensional (3D) scene that includes a 3D model simulating a gesture that represents a letter, a word, or a phrase in a sign language. The method includes obtaining a value indicative of a total number of training images to be generated, using the value indicative of the total number of training images to determine a plurality of variations of the 3D scene for generating of the training images, applying each of plurality of variations to the 3D scene to produce a plurality of modified 3D scenes, and capturing an image of each of the plurality of modified 3D scenes to form the training images for a neural network of the sign language translation system.

The invention relates to the technical field of behavior feature recognition, in particular to a pattern recognition system for fish abnormal behaviors in an underwater video, and the system comprises a fish body skeleton positioning module which is used for retrieving a fish body pixel region in an underwater video frame, positioning a head, a plurality of spine midpoints and a tail fin base, and obtaining a fish body coordinate point set. According to the method, frame-by-frame positioning and key point extraction are carried out on a fish body pixel region in an underwater video frame, a continuous skeleton chain is formed by coordinates of a head, a spine midpoint and a tail fin base part, and a skeleton dynamic sequence with a time attribute is generated based on spatial position information, so that reconstruction of an overall motion structure of a fish body is realized; and further, analysis processing is carried out on the connecting line angle between the skeleton key points, the angle change rate and the skeleton curvature, so that the kinematics description of the dynamic structure of the fish body is refined, meanwhile, complete attitude time sequence information is constructed, and the recognition capability of local deformation is improved.

The invention discloses a transformer substation temperature and humidity sensor detection method and system, and relates to the technical field of transformer substation intelligent monitoring, and the method comprises the following steps: installing a temperature and humidity sensor according to a detection demand, and collecting data information; carrying out statistical analysis on the collected data information to obtain correlation coefficients between temperature and humidity of different areas; introducing a time sequence analysis model, analyzing the long-term trend and short-term fluctuation of the temperature and humidity, drawing a trend chart, and obtaining a periodic mode in the temperature and humidity data; key features are extracted from a periodic mode in the temperature and humidity data, and feature dimensions are reduced; a detection interface is generated, temperature and humidity threshold values are set, and when the temperature and humidity exceed the temperature and humidity threshold values, an alarm system is given out. According to the invention, key features in the temperature and humidity data are extracted and pattern recognition is carried out, so that the system can automatically classify a normal pattern, an early warning pattern and an abnormal pattern, can timely obtain abnormal conditions of the temperature and humidity of the transformer substation, and carries out intervention and processing in advance.

The invention provides an evolution mode recognition method and system based on time sequence tensor segmentation, and the method comprises the steps: collecting three-order time sequence tensor data, and taking the three-order time sequence tensor data as input data; dividing the third-order time sequence tensor into a plurality of tensor segments through a sliding window; extracting a core tensor of the tensor segment based on Tuck decomposition, and performing clustering based on a time sequence weighted clustering method to obtain a coarse-grained segmentation result; optimizing the coarse granularity segmentation result based on information gain to obtain a fine granularity segmentation result; extracting correlation characteristics of each fragment based on the maximum information coefficient; and introducing a neighbor rule, and identifying an evolution mode corresponding to each fragment by adopting an unsupervised three-way clustering algorithm. The invention further constructs an evolution mode recognition system based on time sequence tensor segmentation, a data acquisition part, a time sequence tensor segmentation and evolution mode mining part and a monitoring APP part are connected with one another to form a complete system, and effective recognition of segmentation of the time sequence tensor and the evolution mode is achieved.

The invention relates to the technical field of elevators, in particular to an automatic monitoring and alarming system for elevator traction force, and solves the problems that in the prior art, due to the complexity of the elevator operation environment, the automatic monitoring and alarming system for the traction force may be disturbed, the monitoring precision of the automatic monitoring and alarming system for the traction force may still be limited to a certain extent, and the elevator traction force cannot be monitored. Therefore, the system can not accurately judge whether the traction force is abnormal or not, so that the reliability of the system is influenced. An automatic monitoring and alarming system for traction force of an elevator comprises data acquisition, data processing and analysis, anomaly detection, alarm triggering, information transmission and recording, emergency response and fault processing. According to the system, the change of the traction force can be monitored in real time, an alarm is rapidly triggered under the abnormal condition, elevator accidents such as falling and top rushing caused by insufficient or overlarge traction force are effectively prevented, the safety of elevator operation is remarkably improved, and the system can find potential progressive faults in advance through trend analysis and fault mode recognition.

This invention relates to a multi-task motion pattern recognition system and method for underwater exoskeleton robots. Through innovative data preprocessing and network architecture design, it utilizes a small amount of IMU data to synchronously, in real-time, and with high precision complete motion pattern classification and continuous motion phase prediction, thereby driving the exoskeleton to achieve precise and synchronous assisted control. The multi-task motion pattern recognition system for underwater exoskeleton robots includes a sensing module for acquiring raw posture data of a diver; a data processing module for receiving and calculating relative IMU posture data within the human coordinate system to construct an input tensor; an intent recognition module, including a multi-task deep neural network based on an inverted Transformer architecture, for receiving the input tensor and synchronously outputting motion pattern classification results and continuous motion phase prediction results; and a control module for generating control signals for the underwater exoskeleton based on the motion pattern classification results and continuous motion phase prediction results.

The application relates to the technical field of information technology of construction engineering, and particularly discloses a bidding document review auxiliary method and system based on artificial intelligence, which comprises bidding document preprocessing, technical specification similarity analysis, commercial specification consistency analysis and analysis result collection. The scheme establishes an objective quantitative evaluation system based on the combination of TF-IDF statistical characteristics and pre-training word vector semantic characteristics, provides accurate numerical similarity scores, enhances the robustness of technical specification text content detection through deep semantic understanding, realizes integrated data analysis through a unified numerical extraction framework, can find cross-table type correlation abnormalities, adopts adaptive extraction technology based on numerical pattern recognition, constructs a multi-dimensional abnormal pattern recognition system, effectively detects bid-rigging behavior among bidders, introduces feature contribution degree analysis and abnormal pattern recognition mechanism, and greatly improves the reliability of analysis results.

An example method of training a pattern recognition system. The method may include inputting a voltage based on a pattern to a first set of rows and grounding a first column of the memristor columns via a resistor. The method may also include applying a first scaled voltage to a second column of the column pair, a second scaled voltage to remaining columns, and a third scaled voltage to a second set of rows different from the first set of rows.

The present invention relates to the technical field of traffic pattern recognition, and specifically to an urban traffic pattern recognition system based on trajectory clustering. The system includes a trajectory credibility elimination module, a trajectory semantic clustering module, a trajectory evolution stage recognition module, a dynamic stop determination module, and an urban traffic pattern generation module. In the present invention, by performing boundary discretization rate difference analysis on traffic trajectory data and identifying continuous cross-grid anomalies, unstable segments in the trajectories are effectively excluded. Based on the concentration degree of spatial direction distribution, by calculating the difference in trajectory coverage rate of the clustering center in adjacent time periods, the spatial evolution trend of the trajectory is quantified, realizing the dynamic characterization of traffic patterns in the time dimension. Combining the dual indicators of spatial overlap degree and stop duration between trajectory segments, the risk of misjudgment caused by a single time or space indicator is avoided. Through the similarity matching between trajectory features and travel features, the efficient recognition of the travel function types of urban areas is achieved.

The invention relates to a method for pattern recognition to assemble printed circuit boards by digital printing and a system for implementing the method. In the method, pattern recognition techniques are applied to extract information from imaging of the printed circuit board, thereby enabling automated printing of a targeted object on the board. The method comprises digital printing of a material that is uniformly coated on a transparent donor substrate and positioned in parallel to a receiver substrate. A laser pulse irradiates the donor substrate, resulting in detachment and transfer of a portion of the material onto the receiver substrate. The method enables reduction of overall processing time while achieving high alignment accuracy and high printing resolution.

The invention relates to the technical field of artistic pattern recognition, and provides an artistic pattern recognition system based on artificial intelligence, which comprises a data acquisition module, a feature extraction module, a feature selection module, a texture analysis module and a result evaluation module.The artistic pattern recognition system overcomes the defects in the prior art and is reasonable in design and high in practicability. A digital image of the surface of an artwork is obtained through high-resolution digital imaging equipment, denoising, white balance correction and image enhancement processing are carried out to ensure image quality, then texture feature vectors are extracted through a rotation invariant local binary pattern method, texture features can be effectively extracted, rotation invariance is achieved, and classification robustness is improved; the screened features are input into the pre-trained deep learning model, the artistic pattern category probability is output, the deep learning model is used for learning texture features, and the generalization ability and classification performance of the model are improved.

A chart pattern recognition method includes pre-processing a chart pattern group to be compared to obtain the first chart pattern data to be compared based on time and space; normalizing the first chart pattern data to be compared and the first historical chart pattern data from a database to obtain the second chart pattern data to be compared and the second historical chart pattern data; comparing the second chart pattern data to be compared with the second historical chart pattern data to obtain multiple similarity scores of the second chart pattern data to be compared and the second historical chart pattern data; sorting the multiple similarity scores, the corresponding second chart pattern data to be compared, and the corresponding first historical chart pattern data; and obtaining first historical chart pattern data with a maximum similarity score.

The present invention provides a method for rapid and non-destructive authenticity identification of agarwood based on microscopic imaging and an electronic nose. The method is based on a microscopic image-based agarwood authenticity identification model and an odor-based agarwood authenticity identification model, and comprises the following steps: S1: obtaining a microscopic structural image of an agarwood test sample using a microscope and an imaging system; S2: inputting the microscopic structural image of the agarwood test sample into the microscopic image-based agarwood authenticity identification model; S3: obtaining data of the agarwood test sample in each sensor of the electronic nose according to the odor-based agarwood authenticity identification model; S4: activating the automated pattern recognition system of the electronic nose to perform electronic nose odor recognition data for 40 seconds, 50 seconds, and 60 seconds of agarwood testing, respectively. The method of the present invention is simple, efficient, easy to operate, and does not require sample pre-treatment. It also overcomes the problems of subjective influence and poor repeatability exhibited in traditional manual evaluation, thereby greatly improving the accuracy of authenticity identification of agarwood.

The invention discloses an ERT-based two-phase flow pattern recognition system and a heat transfer coefficient model construction method. The ERT-based two-phase flow pattern recognition system comprises a refrigerant liquid storage tank, a semiconductor chilling plate, a heating diaphragm, a transparent polytetrafluoroethylene tube, a subcooler, a gear pump, a buffer, a filter, a flowmeter, a preheating section inlet pressure sensor, a preheating section inlet thermocouple, a preheating section, an inlet copper tube, an experiment section inlet thermocouple, an experiment section and an experiment section thermocouple. The device comprises an experiment section inlet pressure sensor, a pressure difference sensor, an outlet copper pipe, an experiment section outlet thermocouple, a signal transmission cable, a data acquisition and processing unit, an image reconstruction unit, a post-processing computer, a condenser, an ice-water mixing barrel, a water pipe, a water pump and a three-way valve. The flow pattern distribution condition in the opaque copper pipe can be monitored and displayed in real time, the displayed flow pattern image has high resolution, and the heat transfer coefficient prediction method constructed based on the flow pattern in the opaque copper pipe is simple and high in applicability.

Methods, apparatuses and systems directed to pattern identification and pattern recognition. In some particular implementations, the invention provides a flexible pattern recognition platform including pattern recognition engines that can be dynamically adjusted to implement specific pattern recognition configurations for individual pattern recognition applications. In some implementations, the present invention also provides for a partition configuration where knowledge elements can be grouped and pattern recognition operations can be individually configured and arranged to allow for multi-level pattern recognition schemes.

Systems and methods are provided for determining values regarding a system, such as a vehicle, to enable control of the system. A method includes receiving an indication that a transmission of a vehicle is in a continuous single setting; receiving an indication of a braking event of the vehicle; determining a first energy value for the vehicle based on the transmission being in the continuous single setting and the indication of the braking event; receiving an indication of a change in operation of at least one of the transmission and the braking event; determining a second energy value for the vehicle based on the indication of the change in operation of the at least one of the transmission and the braking event; determining a value for the vehicle based on the first energy value and the second energy value; and controlling operation of a component of the vehicle based on the determined value.

This invention discloses a method for recognizing loading and unloading pallet status patterns. First, a pallet status dataset is constructed. Based on camera positions, the pallet status is divided into five categories and labeled. Next, a ResNet50 network model is constructed, adding ReLU, dropout, and logSoftmax regression layers to adapt to the dataset for sample training. After model convergence, a pallet status pattern recognition system is built. The video stream is input into the recognition system using the RTSP protocol to complete the initial recognition of the pallet status. Median filtering is used to denoise and improve the recognition accuracy, ultimately achieving an accuracy of 99%. This solves the problem of mismatch between work time and vehicle number caused by misjudgments in traditional algorithms. The license plate area is located using the YOLOv5 algorithm, and the target text is extracted using a text detection algorithm, combined with a convolutional recurrent neural network to recognize the license plate information. Finally, the recorded pallet vehicle docking and exit times and work duration data provide a basis for pallet management. This invention realizes the status recognition of each pallet, records the pallet operation status, and can effectively provide feedback on abnormal operations, facilitating management and review.

The application discloses a big data highway accident mode recognition system, comprising: a multi-source data acquisition module for collecting environmental data, facility operation data and vehicle behavior data of a highway section; a data processing module connected with the multi-source data acquisition module, used for fusing and processing various collected data to generate a spatio-temporal aligned joint data set; a mode recognition module connected with the data processing module, used for extracting and fusing multi-dimensional features from the joint data set, and identifying the cause mode and precursor mode of the accident based on the features; and an early warning output module connected with the mode recognition module, used for issuing a road section risk early warning information; the application integrates three heterogeneous data sources of environment, facility and vehicle behavior, constructs a multi-dimensional three-dimensional risk perception system, significantly improves the discovery ability and recognition accuracy of implicit or early risks such as thin ice and fog, and realizes the leap from single appearance monitoring to multi-dimensional essential perception of highway risks.

The invention provides a rare immune fixation electrophoretogram identification system based on a two-stage classification normal form, which is not limited to classification of common disease types, but adds identification of rare disease types on the premise of ensuring common disease type identification precision, so that the identification accuracy of the common disease types is improved. The vacancy of an existing immunofixation electrophoretogram recognition method based on deep learning in the aspect of rare disease type recognition is made up, and universality and practicability are greatly enhanced; meanwhile, a double-stage classification normal form method is adopted, so that the class imbalance ratio of the data set is effectively improved, and the influence caused by class imbalance of the data set is reduced; finally, a loss function LMF specially used for solving data class imbalance in the medical field is adopted, the recognition precision of a few classes is improved on the premise that the recognition precision of most classes is guaranteed, and then the overall recognition precision is improved.

A fault mode recognition method and system based on an adaptive sliding window, according to different fault modes of mechanical equipment corresponding data anomaly characteristics, through Fourier transform, frequency domain information of a kind of mechanical equipment test data is obtained, mechanical equipment test data containing normal state and various fault labels is used for training, the characteristic frequency band of various faults is locked and the data characteristics of various faults are learned, so as to construct a mechanical equipment fault mode discrimination method, through the fault mode recognition system and method, the rapid discrimination of the fault mode of the equipment can be realized, and the characteristic frequency band information corresponding to various fault modes is provided for subsequent research.

The application relates to the technical field of phase pattern recognition, and specifically discloses a PRPD phase pattern recognition system for partial discharge of a distribution network cable based on deep learning. The system comprises: a feature extraction module which extracts multi-scale features to obtain an effective prediction frame; a subordinate screening module which performs spatial overlap analysis, divides the prediction frame which is interfered into a dominant scale frame and a subordinate scale frame; a feature analysis module which calculates the phase domain distribution gradient of the subordinate scale frame and quantifies an independent saliency; a condition construction module which dynamically relaxes the global filtering criterion according to the saliency and generates a local reservation condition; and a constraint verification module which introduces the local reservation condition for deduplication verification, and outputs the final defect category and positioning result. The application is beneficial to isolating independent real defects and false associated noise under the condition of spatial nesting and overlapping of multiple discharge sources, and improves the positioning recognition accuracy and engineering reliability under the condition of multiple defect concurrency.

Systems, methods, and non-transitory computer readable media including instructions for implementing a runtime virtual barrier for fine grained execution control are disclose. Implementing the runtime virtual barrier for fine grained execution control includes receiving, by an application capable of JavaScript execution, an executable code including an API invocation; intercepting, by a virtual barrier, the API invocation; determining that the API invocation is an invocation for a native API configured for subsequent execution in response to a trigger event; based on the determination that the API invocation is an invocation for a native API configured for subsequent execution, recording an invocation source identifier; and upon occurrence of the trigger event: retrieving the invocation source identifier; and influencing execution of the native API based on the invocation source identifier.

The present disclosure relates to a system and method for Al-driven intelligent contact management, privacy enforcement, and dynamic communication control in mobile devices. The system comprises an Al-powered processing unit configured to analyze contact usage patterns, categorize contacts into adaptive groups, and automate the deletion of temporary or redundant entries. A Pattern Recognition System detects frequent, inactive, or spam contacts, while a Blockchain Security Module ensures tamper-proof access logs and secure app permission management. The AI-Based Privacy Enforcement System dynamically restricts unauthorized access to contact data, enforces user-defined privacy settings, and integrates with selective Silent Mode, Do Not Disturb (DND), and call filtering configurations. The system further enables loT-based contact synchronization, ensuring seamless multi-device privacy control and communication optimization. By integrating machine learning, predictive analytics, and decentralized security, the present invention enhances privacy, security, and efficiency in mobile contact management.

The present application relates to cell injection instrument motion control technical field, especially based on the cell injection instrument motion intelligent control system of artificial intelligence, its technical scheme includes: through the construction multimodal perception fusion module and cell mechanics characteristic on-line identification module, fundamentally solved the problem of poor universality and high risk of injury caused by biological sample diversity; The introduction of adaptive motion planning and decision-making based on reinforcement learning center, the motion control problem is converted into a continuous learning optimization process; The designed feedforward and feedback composite execution module, especially its built-in iterative learning controller, effectively solves the bottleneck that the repeatability error is difficult to eliminate in the precision motion system; The integrated abnormal state autonomous diagnosis and recovery module gives the system high robustness and fault tolerance ability; Through the parallel real-time pattern recognition, the system can quickly and accurately diagnose a variety of common operation faults, and trigger intelligent recovery strategy, greatly reduce the dependence on artificial intervention.

The application discloses a GIS internal intermittent discharge mode recognition system and method based on SSA and RF, acquires a PRPD two-dimensional graph of a GIS internal intermittent discharge signal and extracts characteristic parameters thereof; performs dimension reduction processing on the characteristic parameters, extracts principal components with contribution degrees higher than a set threshold value to form a characteristic vector group; finds optimal parameters by using a sparrow search algorithm to obtain a random forest optimized by the sparrow search algorithm; trains a random forest model by using a characteristic vector group of a training sample set, constructs an optimal GIS intermittent discharge fault diagnosis model, classifies a test sample set, and outputs a classification result of the test sample set. The application optimizes learning parameters in a random forest algorithm by using a sparrow search algorithm, solves the problem that parameters of a traditional random forest algorithm are difficult to select, and can more accurately and efficiently recognize GIS internal intermittent discharge fault types.

The invention belongs to the technical field of electronic reconnaissance, and relates to a radar working mode recognition system based on the Bayesian theory and a training method thereof. The radar working mode recognition system comprises a data preprocessing module and a deep neural network module based on Bayesian reasoning. A deep neural network module based on Bayesian reasoning performs Bayesian reasoning on the prediction subset based on the prior subset to obtain a high-dimensional reasoning matrix of the prediction subset; according to the Bayesian theory, real posterior prediction distribution of the prediction subset is obtained based on the priori subset; probabilistically converting the high-dimensional reasoning matrix into posterior prediction distribution approximate to a neural network; and training the neural network to obtain an optimal parameter, so that the approximate posterior prediction distribution of the neural network is approximate to the real posterior prediction distribution, and carrying out radar working mode recognition by using the deep neural network with the optimal parameter. According to the method, the problems of pulse random deletion and prior deletion can be effectively solved, and the working mode recognition capability of pulse random deletion and prior unknown pulse sequences is improved.