57 results about "Student training" patented technology

The invention provides a course reconstruction method and system based on a knowledge graph, and relates to the technical field of intelligent education. The method comprises the following steps: collecting course resource data and student learning behavior data; constructing a knowledge graph fusing multiple courses; generating a personalized learning path based on the atlas and the behavior data; arranging experiment training tasks according to a path and collecting feedback data to update a graph state; and converting the student training result into a technical manuscript in a structured manner and accessing the technical manuscript to a management platform. Closed-loop optimization of teaching resource organization, path recommendation and achievement management is realized.

The invention discloses a model training method and device, electronic equipment and a storage medium. Comprising the following steps: sampling an original video frame sequence based on a target training temporal-spatial resolution to obtain target training data; inputting the target training data into a teacher model, generating a target token set through dynamic token selection, and generating teacher training features through forward propagation; performing multi-scale cutting on the target token set according to a target self-attention weight of the teacher model to generate at least three student training masks with different token numbers; inputting the target training data and the different student training masks into a student model for forward propagation, and generating student training features; and performing alignment distillation on the student training features and the teacher training features to obtain a target student model. The defect of a video understanding model in downstream flexible reasoning is overcome, and dynamic token selection and multi-scale mask training under high temporal-spatial resolution are utilized, so that the model can obtain better performance under various downstream calculated amount limits.

Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for using a large language model (LLM) to generate a specificity labeled query-to-document scored teacher training set for training a teacher model. The teacher model, in turn, in sued to generate a specificity labeled query-to-document scored student training set for training a student model. The student model is ensemble trained, using the student model specificity training set and at least one other student model training set does not quantify the query-to-landing page specificity drift, to predict a performance of a query for a landing page.

The invention relates to the technical field of power transmission line defects, and discloses a power transmission line defect detection method and system based on knowledge distillation, and the method comprises the steps: obtaining a to-be-detected image of a power transmission line; the to-be-detected image is input into a pre-constructed defect detection model, a defect detection result of the power transmission line is obtained, the defect detection model is constructed based on a convolutional neural network, and a teacher-student training method based on knowledge distillation is adopted for training to obtain a defect detection result of the power transmission line; the knowledge distillation comprises any one or more of shallow feature knowledge distillation, deep feature knowledge distillation and response knowledge distillation. According to the method, a multi-level feature distillation and knowledge transfer mechanism is introduced between the teacher model and the student model, multi-level knowledge migration from the teacher model to the student model is realized, neglect of key details caused by simplification of the student model is effectively avoided, the calculation complexity of the student model is reduced, and meanwhile, the method has the advantages of high detection precision and high reliability. Therefore, the overall efficiency of power transmission line defect detection is obviously improved.

Provided is an approach for knowledge distillation based on exporting Principal Components approximations (e.g., Bregman representations) of one or more layer-wise representations of the teacher model. In particular, the present disclosure provides an extension to the original Bregman PCA formulation by incorporating a mean vector and orthonormalizing the principal directions with respect to the geometry of the local convex function around the mean. This extended formulation allows viewing the learned representation as a dense layer, thus casting the problem as learning the linear coefficients of the compressed examples, as the input to this layer, by the student network. Example empirical data indicates that example implementations of the approach improve performance when compared to typical teacher-student training using soft labels.

The invention discloses a big data and artificial intelligence teaching practical training platform based on a computer, and relates to the field of practical training teaching, and the platform comprises a teaching unit which is used for teaching students; a teaching interaction unit which is connected with the teaching unit, can provide an experiment scene for students through the teaching interaction unit, and is used for consolidating the learning content of the teaching unit; the data processing unit is connected with the teaching interaction unit and is used for recording the practical training data of the students and analyzing the recorded data; according to the big data and artificial intelligence teaching training platform based on the computer, learning tasks completed by students can be evaluated, the knowledge mastering and applying degree of the students can be known, and meanwhile, according to the performance, including the cooperation ability, the contribution degree, the communication skill and the like, of the students in group cooperation, the knowledge mastering and applying degree of the students can be known. And evaluation and subsequent learning progress are given.

Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for using a large language model (LLM) to generate a specificity labeled query-to-document scored teacher training set for training a teacher model. The teacher model, in turn, in sued to generate a specificity labeled query-to-document scored student training set for training a student model. The student model is ensemble trained, using the student model specificity training set and at least one other student model training set does not quantify the query-to-landing page specificity drift, to predict a performance of a query for a landing page.

The invention relates to the technical field of intelligent teaching, and provides an engineering integrated intelligent teaching system for rail transit intelligent operation and maintenance. In the system, a teaching data center module aggregates and processes teaching data; the teaching twinborn body module constructs a three-layer twinborn model and generates a multi-dimensional teaching task; the engineering and learning integrated teaching engine module automatically arranges an optimal teaching activity sequence through reinforcement learning according to student ability portraits and teaching tasks; the element universe scene construction module constructs an immersive simulation training scene according to the sequence; the ability evolution evaluation module analyzes trainee training data and generates an ability increment graph; and the teaching feedback optimization module feeds back the atlas to the task generator to drive a new round of teaching. According to the invention, teaching dynamic generation, intelligent arrangement, accurate evaluation and closed-loop optimization in the field of rail transit intelligent operation and maintenance are realized, and the problems of insufficient intelligence, mismatching of teaching strategies, mismatching of engineering and single teaching quality evaluation in rail transit teaching are solved.

The invention relates to the technical field of intelligent training and data analysis, in particular to an intelligent model construction method and system for student training data analysis. Comprising the following steps: constructing an SOP time sequence knowledge graph, obtaining a standard operation specification document and expert-level operation demonstration data, and constructing the SOP time sequence knowledge graph; real-time feature extraction: capturing a multi-modal data stream, and generating a real-time behavior feature vector; state recognition and path deduction: inputting the real-time behavior characteristics into an event classifier to recognize a key event, performing path deduction based on deterministic finite state automaton logic in an SOP time sequence knowledge graph, and determining and outputting a current state pointer; deviation analysis and judgment: extracting a compliance mask based on a current state, and injecting the compliance mask into a decoding process of the attention network; and performing feedback and adjustment to generate a corresponding deviation event signal. Through deep analysis, feedback is more targeted, and the accuracy and fineness of training evaluation are remarkably improved.

The invention relates to the technical field of training teaching, in particular to an exercise training action teaching system. The system comprises a training data acquisition module, a trainee action difference analysis module, a trainee training teaching module and a demonstration teaching feedback adjustment module, and can acquire corresponding training action postures and acceleration data in real time by configuring an inertial sensor on sports equipment corresponding to a trainee; meanwhile, the corresponding body consumption load condition is obtained; performing student action difference feature analysis on the training action postures and the acceleration data based on the body consumption load condition to obtain training action difference features corresponding to different students in the daily exercise training process; training teaching resources and time limiting conditions are obtained, and student training teaching analysis and teaching feedback adjustment are carried out, so that an exercise training teaching feedback adjustment strategy corresponding to the student is generated. According to the invention, scientific and personalized action teaching services can be provided for exercise training learners, so that the exercise training effect is improved.

The invention relates to the technical field of acupuncture training models, and particularly discloses a traditional Chinese medicine acupuncture training model which comprises an arm model. Puncturing the epidermis; the muscle simulation mechanism is located below the puncture epidermis; an acupuncture depth detection mechanism; according to the traditional Chinese medicine acupuncture and moxibustion training model, muscles with different compaction degrees are simulated through the arranged muscle simulation mechanism, so that the influence of the muscles with different compaction degrees on filiform needle puncture during acupuncture and moxibustion of patients with different constitutions such as obesity or strong constitutions in reality is simulated; a medical student is assisted in training finger force, the acupuncture depth and the acupuncture angle of the medical student during acupuncture are detected through the arranged acupuncture depth detection mechanism, and the air feeling simulation mechanism simulates muscle contraction reaction when the medical student controls the acupuncture needle to lift, insert and twist, so that the medical student is assisted in experiencing air feeling during acupuncture. Therefore, medical students can practice acupuncture and moxibustion under the condition close to reality, and the acupuncture and moxibustion skill is improved.

The invention discloses a cognitive diagnosis method based on federated learning, which is used for jointly training a cognitive diagnosis model based on a personalized strategy of parameter decoupling, protecting data privacy and realizing accurate and fair diagnosis of students of each client. A local cognitive diagnosis model is trained for students in each client according to local student learning data, each client uploads part of model parameters to a central server for parameter aggregation according to a designed personalized strategy based on parameter decoupling, and federal learning and cognitive diagnosis are sought to be combined. The method comprises the following steps of: firstly, updating shared parameters through difference perception aggregation and loss perception aggregation to ensure the diagnosis accuracy of a cognitive diagnosis model and protect the privacy of a user, updating the shared parameters through difference perception aggregation and loss perception aggregation on the basis of a personalized strategy of parameter decoupling, and locally storing the parameters of a diagnosis function as private parameters.

Method for training a student network of a teacher-student training architecture for point cloud registration for mapping an environment for navigating a technical system, wherein the teacher-student training architecture comprises a teacher network (200) and the student network (202).

The utility model discloses a multifunctional monitoring device for trainee training, comprising a monitoring display which is connected with a cardiopulmonary electrode patch, a limb monitor and a brain monitor through connecting wires and is used for multifunctional monitoring of trainee training; the cardiopulmonary electrode patch comprises a pasting piece, a first fixing shell and a monitoring electrode, the bottom of the first fixing shell is fixedly connected with the circular pasting piece, the monitoring electrode is clamped in the first fixing shell, and the monitoring electrode is used for training and monitoring of trainees. A cardiopulmonary electrode patch, a limb monitor and a brain monitor are arranged on the multifunctional monitoring device, a first monitoring electrode is fixedly installed on a first fixing shell on the cardiopulmonary electrode patch and used for training cardiopulmonary monitoring of a trainee, and meanwhile a second monitoring electrode is clamped to a second fixing shell on the limb monitor and used for monitoring the heart and the lung of the trainee. Therefore, the second monitoring electrodes detect the limbs of the trainee.

An end-to-end automatic speech recognition (ASR) system can be constructed by fusing a first ASR model with a transformer. The input of the transformer is a learned layer generated by the first ASR model. The fused ASR model and transformer can be treated as a single end-to-end model and trained as a single model. In some embodiments, the end-to-end speech recognition system can be trained using a teacher-student training technique by selectively truncating portions of the first ASR model and / or the transformer components and selectively freezing various layers during the training passes.

The application discloses a kind of training fatigue accumulation warning method and system, it is related to digital training warning technical field, it is difficult to identify the safety risk of fatigue accumulation of student in the process of training Problem.The application embodiment is in the case where the time consumed by student training exceeds time consumption baseline, based on the deviation between timeout amplitude, physiological data and physiological baseline, and the fatigue conduction coefficient between training nodes Calculate cumulative fatigue index, the risk brought by the fatigue accumulated by student can be measured;Then calculate training error to measure the conversion tendency of the risk, and issue a call signal based on the conversion tendency to alarm.Using the application embodiment, the safety risk of fatigue accumulation of student in the training process can be accurately identified, and timely alarm.

The utility model relates to the technical field of medical puncture training, in particular to puncture training equipment based on a delta parallel mechanism, which comprises a base and a base, and a support is arranged on the front side of the base. Comprising a delta parallel mechanism, the delta parallel mechanism is provided with a driving end and a polymerization end, and a puncture needle is connected to the polymerization end in a penetrating mode; the device further comprises a torque generating mechanism, the torque generating mechanism comprises a motor, a semicircular force arm and a wire, the wire is connected between the output end of the motor and the semicircular force arm, and power is transmitted through wire driving. The motor is assembled on the support, the semicircular force arm is hinged to the support, and the driving end of the delta parallel mechanism is hinged to the semicircular force arm. Force feedback can be accurately simulated in the puncture training process, a more real puncture training effect is provided, high application precision is kept in the repeated use process, and the device is suitable for medical student training and has an extremely good market popularization prospect.

Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for generating a teacher training set of specifically tagged query-to-document scores for training a teacher model using a large language model (LLM). The teacher model is in turn used to generate a student training set of specifically tagged query document scores for training student models. The student model is integrally trained to predict performance of the query for the login page using the student model-specific training set and at least one other student model training set that does not quantify the query to the login page-specific deviation.

A system, method and computer program product for training a deep neural network model using a pre-trained teacher model. Student training data samples are input to the deep neural network model and teacher training data samples are input to the pre-trained teacher model. The trained deep neural network model is generated using the training data samples to optimize an error function that is evaluated using a plurality of student label prediction outputs and a plurality of Markov transformed teacher label prediction outputs. Each Markov transformed teacher label prediction output is generated based on a teacher label prediction output by the pre-trained teacher model in response to receiving one of the teacher training data samples as an input. Each Markov transformed teacher label prediction output is generated through a Markov transform involving matrix multiplication using a Markov matrix.

The invention relates to the technical field of aviation safety and police service practical training evaluation, and discloses an air police new quality combat capacity civil aviation safety practical training evaluation method. A demand information acquisition module, a course supervision module, a teacher team module, a basic ability analysis module, an actual combat simulation ability analysis module, a comprehensive ability evaluation module, a feedback optimization module and a result solidification module are arranged in the model, and the demand information acquisition module performs multi-dimensional information acquisition; the course supervision module performs course quality and suitability supervision; the teacher team module executes selection of teacher teams and continuous ability improvement; the basic capability analysis module executes student basic quality evaluation; the actual combat simulation capability analysis module executes refined evaluation of scenario actual combat performance; the comprehensive ability evaluation module executes the end evaluation of the practical training effect of the trainee; the feedback optimization module outputs targeted optimization suggestions; and the achievement solidification module establishes a normalized graduate tracking return visit mechanism.

The invention belongs to the technical field of medical image segmentation, and particularly relates to a semi-supervised medical image segmentation method based on a sequence convolution relation. The method comprises the steps of firstly performing data preprocessing on an original medical image to obtain medical images with the same size, the original medical image including an unlabeled medical image and a labeled medical image; performing copy-paste operation on the preprocessed medical image without the label and the preprocessed medical image with the label to generate a mixed image; constructing a teacher-student training framework, and inputting the mixed image into a student network model to generate a segmentation mask; inputting the preprocessed medical image without the label into the teacher network model to generate a pseudo label, and performing copying-pasting operation on the preprocessed medical image without the label and the preprocessed medical image with the label again to generate a new mixed image; and carrying out loss calculation on the new mixed image and the segmentation mask, and stopping iteration to obtain a final parameter model until an iteration number threshold is met, thereby determining a final image segmentation model.

The invention discloses a teaching evaluation method and system based on multi-modal data fusion, and a storage medium, and the method comprises the following steps: obtaining multi-modal data of students, the multi-modal data comprising texts, images, voices and behavior data generated in the learning process of the students; fusing the multi-modal data to form a unified representation; carrying out deep learning analysis on the fused multi-modal data, and carrying out deep modeling on the multi-modal data; analyzing the modeled multi-modal data to generate a student training scheme; and sending the student training scheme to the teacher and student ports. According to the invention, through multi-modal data acquisition, intelligent preprocessing, fusion, deep learning analysis, intelligent evaluation and analysis, dynamic recheck and teaching data feedback, in combination with data acquisition, processing, storage and transmission modules and an output interface, efficient processing and application of teaching data are realized, and the teaching system is optimized.

The application relates to a cardiopulmonary resuscitation immersive training device and method, wherein a simulation man, a host computer and a playing module are in communication with each other, all training operations of a user can be sensed by the simulation man and sent to the host computer for analysis, a database of the host computer stores cardiopulmonary resuscitation training data, and the host computer controls the playing module to play corresponding interactive videos. The application establishes an interactive relationship among the simulation man, the host computer, the playing module and students. In this way, the students participating in the training have strong sense of identification, the students can avoid the thought distraction caused by the too unrealistic simulation environment, meanwhile, the device provides real-time feedback, and good interactivity is generated with the students, so that the quality of the student training is improved.

The invention discloses a teaching scheme design system, which is used for solving the problems of single student training mode, high teaching cost and poor teaching quality in the prior art. According to the invention, the learning ability of students and the teaching ability of teachers are represented by using the homework scores of the students, and the learning ability of the students is corrected by using the enrollment scores and attendance rates of the students, so that the representation results are more persuasive. According to the method, the relation curve between the teaching achievement of teaching after the teaching contents are divided and sorted according to different modes and the learning ability and attendance rate of the students is sought through the expected value in statistics, so that the design of the teaching scheme can be started after the recording work of the students is completed. Along with further accumulation of the teaching data and expansion of the acquisition range of the teaching data, the method can be applied to the design of more teaching course schemes, and has the application values of saving teaching cost and cultivating high-quality graduates and teachers.

The invention relates to a method for providing a trained machine learning model for detecting a hands-on / hands-of state (40) of a steering wheel of a vehicle (50), having the steps of: training at least one teacher model (20) using a training data set (10); training a smaller student model (21) using the at least one teacher model (20) in the form of a teacher-student training process, wherein for this purpose, a loss function is used which is composed of a weighted sum of a student loss and a distillation loss, and a weighting factor (λ) is ascertained, while taking into consideration at least one specified criterion (11), for each piece of training data or group of training data of a training data set (10) used in the teacher-student training process; and providing the trained student model (21). The invention additionally relates to a device (1) for providing a trained machine learning model for detecting a hands-on / hands-off state (40) of a steering wheel of a vehicle (50) and to a method and a device for detecting a hands-on / hands-off state (40) of a steering wheel of a vehicle (50).

The invention discloses a student training full-period management system based on artificial intelligence, and relates to the technical field of artificial intelligence, and the system comprises a student state sensing unit which is used for collecting multi-modal data, carrying out the multi-modal fusion analysis, and obtaining the comprehensive learning input degree; the knowledge graph construction unit is used for carrying out correlation degree calculation and establishing connection between the corresponding knowledge entities when the correlation degree is greater than a preset correlation degree threshold value so as to generate a cross-domain knowledge graph; the ability portrait evaluation unit is used for performing dynamic update analysis in combination to obtain the current mastery degree of the students on the knowledge points; and the learning path planning unit is used for carrying out learning readiness analysis and recommendation utility analysis and generating a personalized learning path. According to the invention, through multi-modal perception, cross-domain map construction and dynamic path planning, accurate and closed-loop management of student states, knowledge structures and ability development is realized, and the efficiency and effect of personalized training are significantly improved.

This invention relates to the field of intelligent teaching technology and provides an integrated engineering-learning intelligent teaching system for intelligent operation and maintenance of rail transit. In this system, a teaching data platform module aggregates and processes teaching data; a teaching twin module constructs a three-layer twin model and generates multi-dimensional teaching tasks; an integrated engineering-learning teaching engine module automatically arranges the optimal sequence of teaching activities based on student competency profiles and teaching tasks through reinforcement learning; a metaverse scene construction module constructs immersive simulation training scenarios based on this sequence; a competency evolution assessment module analyzes student training data and generates a competency increment map; and a teaching feedback optimization module feeds this map back to the task generator to drive a new round of teaching. This invention achieves dynamic generation, intelligent arrangement, precise assessment, and closed-loop optimization of teaching in the field of intelligent operation and maintenance of rail transit, solving problems such as insufficient intelligence, mismatched teaching strategies, mismatch between engineering and learning, and a single method of evaluating teaching quality in rail transit teaching.

The invention provides a medical examination test paper generation system and method, and relates to the field of medical student training, and the system comprises a front-end display layer, a business logic layer and a data storage layer which are in telecommunication connection. The front-end display layer comprises an examination question display module, an answer interface module and a result feedback module, and is used for providing a friendly operation interface for students; the business logic layer is used for processing an assessment process, a scoring rule and data analysis; and the data storage layer is used for storing examination questions, student answer data and examination results. The system has the advantages that the examination efficiency and quality are improved, the workload of examination personnel is reduced, the learning enthusiasm and initiative of medical students are enhanced, the professional ability of the medical students is promoted to be improved, more excellent clinicians are cultivated for medical institutions, and the overall medical service level is improved.

The invention discloses a dance movement teaching system and method based on movement perception, and belongs to the technical field of dance teaching. Physiological data of students during dance training are collected, and the dance difficulty is evaluated according to a dance difficulty evaluation strategy; collecting exercise data of the student after the training unit is completed, and evaluating the tidiness of the student during training according to a dance tidiness evaluation strategy; images of the students during dance training are collected in real time, and standard dance conditions of the students during training are evaluated according to the dance posture evaluation model; evaluating a dance teaching comprehensive condition based on the dance difficulty, the dance uniformity and the dance accuracy; judging whether the dance teacher needs to teach again or not based on the comprehensive dance teaching condition; the training data is collected, the dance difficulty, the dance uniformity and the dance accuracy are combined to analyze the teaching condition, the teaching condition can be mastered more accurately, the wrong posture can be corrected in time, and the dance teaching quality and efficiency can be improved.