48results about How to "Solve problems with high computational complexity" patented technology

The invention relates to the technical field of graphic signal processing, and specifically discloses an online distributed multi-task graph filter construction method based on recursive least squares, including steps: S1: constructing a graph filter with variable nodes; S2: constructing a distributed Learning model to track the coefficient vector of the unknown graph filter; S3: Use the online distributed alternating direction multiplier method to solve the distributed learning model; S4: Utilize the coefficient vector obtained in step S3 to update the node constructed in step S1. Changed graph filter. The present invention uses the recursive least squares method to solve the model, which speeds up the estimation speed of the graph filter; in the graph filter model with variable nodes, when the graph filter coefficients are different but similar, considering the multi-task situation, it can promote The cooperation between nodes improves the estimation speed and estimation accuracy; the algorithm is improved into an online method, which solves the problem of large computational complexity.

The present invention proposes a gesture recognition method and device, and electronic equipment, wherein the method includes: collecting a first image including a gesture; acquiring the center of the gesture according to the first image; and determining the control of the gesture on the terminal device according to the center coordinates Range: Get the control object from the control range, and operate the control object according to the control instruction corresponding to the gesture. Through image processing, the center coordinates of gestures can be obtained, and the terminal equipment can be precisely controlled, and the computational complexity is reduced, which solves the existing gesture recognition method that uses spatial algorithms to calculate the motion trajectory of gestures, and then controls terminal equipment. , a problem with high computational complexity.

The invention relates to an incremental trajectory anomaly detection method based on incremental kernel principal component analysis, which belongs to the incremental trajectory anomaly detection method. This method: first perform the initial calculation of the model, and use the traditional Batch KPCA to calculate the initial kernel feature space. Whenever there are M new trajectory data coming, first standardize the M trajectory data; then use Batch KPCA to calculate the new Increase the kernel feature space of the data; calculate the average reconstruction error of the new data and the training data respectively, if the error of the two is greater than the given threshold, execute the subsequent kernel feature space segmentation-merging method to update the kernel feature space; then update The final kernel feature space is projected to extract the principal components; finally, a class of support vector machines is used for unsupervised learning and anomaly detection. Advantages: This method is superior to the traditional kernel principal component analysis method, reduces computational complexity, and improves the efficiency of trajectory anomaly detection.

The invention relates to a log likelihood estimation based novel passive locating and tracking method. A multistation angle-measurement passive-location non-linear measurement equation is subjected to pseudo-linearization processing by the aid of log likelihood estimation, target position is obtained by means of maximum likelihood estimation, a closed-form solution of a target locating algorithm is given, then a 'current' statistical maneuver model and Kalman filtering are utilized to calculate target position, velocity and acceleration online in real time, and real-time accurate positioning and tracking on the target is realized.

The invention relates to a route planning method and system for oil and gas pipelines. First, according to the digital terrain data of the undulating terrain where the oil and gas pipelines are to be built, and the coordinates of the starting point and the end point of the oil and gas pipelines to be built, a simplified strategy of dividing and conquering the search space is adopted to determine the planned construction. The laying area of ​​oil and gas pipelines can effectively compress the data scale of the digital terrain in the laying area, so that the pipeline routing planning is only carried out in the divided and conquered terrain corresponding to the small-scale laying area, which effectively avoids redundant calculations and solves the problem of the amount of digital terrain data. It is a huge, complex and high calculation problem of oil and gas pipeline routing planning. Then, the improved hybrid frog leaping algorithm is used to iteratively solve the problem in the laying area, and the optimal pipeline routing scheme is calculated, thereby avoiding the problems of large engineering volume, high construction cost, and long scheme formulation period caused by the uncertainty of artificial experience. , to achieve efficient and high-quality design of oil and gas pipeline routing schemes.

The invention discloses a multi-input multi-output visible light communication system based on coefficient separation. At a transmitting end, LEDs are divided into two groups. The first group of LEDs transmits the positive integer part of a bipolar pulse amplitude modulation signal, and the second group of LEDs transmits the absolute value of the negative integer part of the bipolar pulse amplitude modulation signal. At a receiving end, a novel receiver with low computing complexity is utilized; the receivers utilizes the characteristic of a transmitted signal; the absolute value of the bipolar pulse amplitude modulation signal is estimated; and then the polarity of the bipolar pulse amplitude modulation signal is estimated; and finally estimation of the bipolar pulse amplitude modulation signal is finished. The multi-input multi-output visible light communication system with separated positive coefficient and negative coefficient according to the invention has advantages of low computing complexity, high performance, etc.

The invention relates to the field of artificial intelligence security, in particular to a black-box attack defense system based on the regularization of the middle layer of the neural network, including a first source model, a second source model and a third source model; The box attack defense method includes S1, inputting pictures into the first source model for white-box attack, outputting the first adversarial sample sequence, S2, inputting the first adversarial sample sequence into the second source model, and outputting the second adversarial sample sequence , S3, inputting the second adversarial sample sequence into the third source model for black-box attack, outputting the third recognition sample sequence, S4, inputting the third recognition sample sequence into the third source model for adversarial training, updating the third source model ; The adversarial samples generated by this algorithm have the characteristics of high transferability to the target model, and can also effectively defend the target model from being attacked through confrontation training.

The embodiment of the present disclosure discloses a method, device, computer equipment, and storage medium for determining a cluster center. The method includes: converting each two-dimensional geographic location information in the geographic location information set to be processed by using geographic location coding technology It is one-dimensional location coding information; a dictionary tree is generated according to each location coding information, a tree node in the dictionary tree corresponds to a set geographic location area, and the geographic location area corresponding to a child node belongs to the parent node of the child node Within the corresponding geographic location area; according to the positional relationship between each tree node and each geographic location information in the dictionary tree, and the quantity value of the geographic location information associated with each tree node, calculate the corresponding The density value of ; according to the density value, at least one cluster center is determined in the geographic location information set. The technical solutions of the embodiments of the present disclosure can reduce the computational complexity of the cluster center in the clustering algorithm.