45results about How to "Fully explore" patented technology

The invention discloses a multi-dimensional IRT-based teachers teaching ability evaluation method. The evaluation method comprises the following steps: a step of information collection before teachersteaching evaluation, a step of collecting the students evaluation of the teachers classroom language teaching ability which is needed for teachers ability evaluation, a step of preparation for the teachers teaching evaluation, a step of classification analysis on teachers teaching ability, and a step of summarizing the teachers teaching evaluation. The beneficial effects of the method are that via multi-dimensional IRT-based teachers teaching ability evaluation method, comprehensive evaluation of the teachers teaching ability can be realized, operation is simple, the teachers teaching abilityis fully explored, influence exerted by human factors can be prevented, veracity of teachers teaching ability evaluation can be ensured, the method can help reduce the evaluation of teachers teaching, multi-faceted evaluation is adopted, the accuracy of the evaluation of teachers teaching ability can be improved, and evaluation of competent teachers can be facilitated.

The invention discloses a method and a device for power spectral parameter detection and analysis of pulse wave. The device comprises a master control module, a pulse wave detection sensor, a filter module, a sampling module, a liquid crystal display module, a serial port communication module, a SD (secure digital) card memory module and an upper computer software analysis system. The pulse wave detection sensor, the filter module, the sampling module, the master control module, the serial port communication module and an upper computer are connected in sequence. The master control module is respectively connected with the liquid crystal display module and the SD card memory module. Human pulse wave signals are converted into analog electrical signals through the pulse wave sensor, the analog electrical signals filtered by the filter module are converted into digital signals through the sampling module, the pulse wave signals are transmitted to the liquid crystal module for displaying after the digital signals are subjected to signal processing by the master control module, and the pulse wave signals are further analyzed through the SD card memory module and upper computer software. The method and the device can be used for accurately detecting the human pulse wave, and the device, which is simple and easy in operation, is provided for power spectral analysis of the pulse wave.

The invention provides an image dimension reduction method of an extreme learning machine based on graph embedding, and belongs to the field of machine learning and data mining. The method comprises the following steps: firstly, selecting an original image sample set, and constructing a sample relation matrix by utilizing inter-sample distances and label information of original image samples; secondly, according to the constructed sample relation matrix, firstly, carrying out random mapping on an input vectorized image sample, and then, learning a feature extraction matrix by minimizing a weighted sample reconstruction error; and finally, performing data dimension reduction on vectorized image data by using the learned feature extraction matrix. The method is short in training time and efficient in data compression, and the compression quality and dimension reduction stability of the data are effectively improved.

The invention relates to a cattle killing machine, and belongs to slaughtering equipment. The cattle killing machine comprises a prisoner tube mechanism, a front push plate, a rear push plate, a cattle rump drive mechanism, a stranglehold mechanism, a cattle head turnover mechanism, a post-slaughter cattle ejecting mechanism, a cavity door start-stop mechanism and a power drive mechanism, wherein the prisoner tube mechanism comprises first and second bases, first and second beams, a top plate, a bottom plate, a left rotary disk, a right rotary disk and a cattle containing groove; one end of the lower part of the front push plate is arranged on the left rotary disk, and the other end of the lower part of the front plate is arranged on the right rotary disk; one end of the lower part of the rear push plate is connected with the left rotary disk, and the other end of the lower part of the rear push plate is connected with the right rotary disk; the cattle rump drive mechanism is arranged on a wall body on one side of the front push plate, and the stranglehold mechanism is arranged on the center position of the first base; the cattle head turnover mechanism is arranged on the stranglehold mechanism, and the post-slaughter cattle ejecting mechanism is arranged on the upper part of one side of the front push plate; the cavity door start-stop mechanism is arranged on the right rotary disk; and the power drive mechanism is arranged on one side of the first base or the second base, the two ends of the first beam are respectively provided with a rear push plate acting cylinder, and the two ends of the second beam are respectively provided with a front push plate action cylinder. The cattle killing machine provided by the invention has a simplified structure, the prisoner tube mechanism is avoided being damaged, and the balance property of a neck clamping force is good.

The invention discloses an Internet sales planning system in the technical field of internet sales planning, which comprises a server, a cloud platform and an e-commerce platform, and is characterized in that the server is connected with the cloud platform through a wireless network; according to the intelligent marketing system based on the Internet big data, through cooperation of the cloud platform and the e-commerce platform and combination of the Internet big data of the cloud platform, the transaction module and the after-sales module, the purchase efficiency of the user can be further improved, the potential customer market is fully explored, and the market share is improved. According to the system, companies, individuals and dealers are connected with client terminals through the Internet platform, and clients can directly inquire information such as product prices from the companies, individuals and dealers through the wireless Internet platform, so that a traditional marketing mode is changed, people can buy favorite products without going out, and meanwhile, the manpower and material resource expenditures of companies, individuals and dealers for marketing products are reduced, and the system is suitable for large-scale popularization.

The invention discloses an indoor scene outline detection method fusing color and depth information and belongs to the technical field of machine vision. The method comprises the key techniques of 1,separating a color image into R (red), G (green) and B (blue), using an image edge detection operator to extract edge information on each channel, and performing weighted fusing to obtain overall results of color image edge detection; 2, using a depth image to extract edge information, correcting false edges due to the loss of three-dimensional information of a scene only during detection of the color image, and enhancing confidence of true edge detection results; 3, constructing a gradient pyramid, and performing multi-scale multi-level fusing on color image edges and depth image edges to obtain edge detection results, and executing edge aggregation to obtain a final outline. The method of the invention has the advantages that gradient information of a depth image is fully excavated, outline detection results of a color image are corrected and enhanced via the outline detection results of the depth image, and a more robust and precise scene outline is obtained accordingly.

The invention relates to a cattle killing machine, and belongs to slaughtering equipment. The cattle killing machine comprises a prisoner tube mechanism, a front push plate, a rear push plate, a cattle rump drive mechanism, a stranglehold mechanism, a cattle head turnover mechanism, a post-slaughter cattle ejecting mechanism, a cavity door start-stop mechanism and a power drive mechanism, wherein the prisoner tube mechanism comprises first and second bases, first and second beams, a top plate, a bottom plate, a left rotary disk, a right rotary disk and a cattle containing groove; one end of the lower part of the front push plate is arranged on the left rotary disk, and the other end of the lower part of the front plate is arranged on the right rotary disk; one end of the lower part of the rear push plate is connected with the left rotary disk, and the other end of the lower part of the rear push plate is connected with the right rotary disk; the cattle rump drive mechanism is arranged on a wall body on one side of the front push plate, and the stranglehold mechanism is arranged on the center position of the first base; the cattle head turnover mechanism is arranged on the stranglehold mechanism, and the post-slaughter cattle ejecting mechanism is arranged on the upper part of one side of the front push plate; the cavity door start-stop mechanism is arranged on the right rotary disk; and the power drive mechanism is arranged on one side of the first base or the second base, the two ends of the first beam are respectively provided with a rear push plate acting cylinder, and the two ends of the second beam are respectively provided with a front push plate action cylinder. The cattle killing machine provided by the invention has a simplified structure, the prisoner tube mechanism is avoided being damaged, and the balance property of a neck clamping force is good.

The invention discloses a 5G network-based cultural tourism resource collection platform and collection method, including a collection terminal that sends cultural tourism resource data to a cloud server in real time through a 5G network; a cloud server that receives data sent by the collection terminal and stores it in the cloud; a cache unit , to download data from the cloud server; the automatic review system acquires data sequentially from the buffer unit, extracts the trajectory information in the data, judges whether the cultural tourism resource corresponding to the data is unknown according to the trajectory information, and sends the unknown cultural tourism resource data Send to the manual classification system; the manual classification system manually classifies unknown cultural tourism resources; the management system summarizes unknown cultural tourism resource data. The invention is used to solve the problems of low collection efficiency and chaotic and complicated data of emerging cultural tourism resources in the prior art, realize the purpose of improving the collection efficiency of cultural tourism resources, and at the same time facilitate the management and control of tourism management departments.

The invention discloses a light field image super-resolution reconstruction method based on frequency domain analysis and deep learning, which first uses optical flow-based rendering and bicubic interpolation to generate an initial high spatial and angular resolution light field image; then adopts Discrete cosine transform is used to perform frequency domain conversion on the generated initial light field image to obtain a frequency component image that characterizes the characteristics of the light field image; based on this, super-resolution reconstruction can be modeled as a frequency restoration problem, so multiple 2D convolutions are constructed neural network to model the initial restoration of each frequency component, and integrate semantic information into the network to enhance the restoration effect; then combine the frequency components after the initial restoration and build a 3D convolutional neural network for fine restoration; finally, use The inverse discrete cosine transform is used to reconstruct all the restored frequency components into the required light field image; the advantage is that it can effectively improve the spatial and angular resolution of the light field image, and can restore texture information and retain angular consistency.

The invention provides a method for identification and analysis of components in cockroaches, comprising the following specific steps: (1) collecting cockroach samples in five growing stages of male adults, female adults, ecdysis adults, 7th-8th instar nymphs and 3nd-4th instar nymphs; (2) preparing the cockroach samples in different growing stages: extracting with 95% ethanol; (3) optimizing GC-MS coupling analysis conditions: optimizing GC-MS analysis conditions of components in cockroaches in different growing stages by selection of sample port temperature, sample size and sample mode; and (4) identifying specific components of cockroaches in each growing stage. The method of the invention has high sensitivity and reliable identification results, and can be used for identification of cockroaches in different growing stages. By the method, research and utilization of components of cockroaches in each growing stage are enhanced.

The invention relates to an environmental detection method and system for archaeology. The environmental detection method comprises the following steps that archaeological excavation is carried out byusing a robot body and a control body; a processor in the robot body or the control body evaluates and divides a passable area and an unpassable area in response to a parallax image collected by a camera device of the robot body; and the processor or the control body releases at least one first-class relay sensor and / or at least one second-class relay sensor based on the unpassable area, and deploys a plurality of third-class relay sensors based on the passable area, wherein the third-class relay sensors are connected mutually in a multi-hop mode and can receive information transmitted by thefirst-class relay sensors and / or the second-class relay sensors.

The present invention provides a deep separable convolutional neural network acceleration method, including: performing deep convolution on the input neuron, and when performing the deep convolution calculation, the same M lines of the C input channel are in the three-dimensional processing unit PE array Independent and parallel computing, to get the output neurons of the same N rows of the C channel, N

The invention belongs to the field of human-machine dialogue, and specifically relates to a knowledge-driven dialogue strategy network optimization method, system, and device, aiming to solve the problem of low utilization rate of training data in the optimization process of the existing dialogue system in the medical field. The method of the present invention is applied to the training of the dialogue system, including: based on the input characteristics of the dialogue samples in the selected field, the self-encoder is optimized by using the collaborative filtering method based on the encoder; the encoder in the self-encoder is used to The input features are mapped to a low-dimensional space; the dialogue policy network is optimized based on a loss function that includes an additional reward value; the additional reward value is inversely proportional to the frequency of state-action pairs in the behavior-action space. The invention can fully explore the state-action space under the condition of limited training data, and improve the data utilization rate.

The invention discloses a behavior recognition method based on feature mapping and multi-layer time interactive attention, and solves the problem that the behavior recognition capability is insufficient due to the fact that time dynamic information modeling is insufficient and the mutual dependency relationship between different frames is ignored in the prior art. The method comprises the following implementation steps: (1) generating a training set; (2) acquiring a depth feature map; (3) constructing a feature mapping matrix; (4) generating a time interaction attention matrix; (5) generating a time interaction attention weighted feature matrix; (6) generating a multi-layer time interactive attention weighted feature matrix; (7) acquiring a feature vector of the video; and (8) carrying out behavior identification on the video. According to the method, the feature mapping matrix is constructed, and multi-layer time interaction attention is provided, so that the accuracy of behavior recognition in the video can be improved.

Belonging to the technical field of biological genetic engineering, the invention in particular relates to a hordeum vulgare calmodulin gene HvCAM1 and salt tolerant application thereof. The hordeum vulgare calmodulin gene HvCAM1 has a nucleotide sequence shown as SEQ ID No:1 and a full-length genome sequence 2124 bp. According to the invention, firstly a hordeum vulgare calmodulin coded gene is obtained by cloning and is named as HvCaM1, and the transcript cDNA full-length nucleotide sequence is shown as SEQ ID No:2, and the full-length cDNA sequence is 728 bp. The invention also utilizes RNAi interference method to silence the HvCAM1 gene successfully in hordeum vulgare Golden Promise, also utilizes the material to verify the negative regulation effect of HvCAM1 gene under salt stress can effectively enhance the salt tolerance of hordeum vulgare. At the same time, overexpression of the HvCAM1 gene in hordeum vulgare Golden Promise cannot enhance the salt tolerance of hordeum vulgare.