CN111239685B - Sound source positioning method based on uniform design and self-organizing feature mapping neural network - Google Patents
Sound source positioning method based on uniform design and self-organizing feature mapping neural network Download PDFInfo
- Publication number
- CN111239685B CN111239685B CN202010021009.6A CN202010021009A CN111239685B CN 111239685 B CN111239685 B CN 111239685B CN 202010021009 A CN202010021009 A CN 202010021009A CN 111239685 B CN111239685 B CN 111239685B
- Authority
- CN
- China
- Prior art keywords
- neural network
- sound source
- self
- feature mapping
- layer
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
- G01S5/00—Position-fixing by co-ordinating two or more direction or position line determinations; Position-fixing by co-ordinating two or more distance determinations
- G01S5/18—Position-fixing by co-ordinating two or more direction or position line determinations; Position-fixing by co-ordinating two or more distance determinations using ultrasonic, sonic, or infrasonic waves
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- General Health & Medical Sciences (AREA)
- General Engineering & Computer Science (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- Artificial Intelligence (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Biomedical Technology (AREA)
- Life Sciences & Earth Sciences (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Health & Medical Sciences (AREA)
- Radar, Positioning & Navigation (AREA)
- Remote Sensing (AREA)
- Stereophonic System (AREA)
Abstract
The invention provides a sound source positioning method based on uniform design and self-organizing feature mapping neural network, and relates to the technical field of sound source positioning. Firstly, determining a three-dimensional space where a sound source is located, and dividing the three-dimensional space into n three-dimensional grids; selecting m characteristics of the sound signal as experimental factors in a uniform design method; selecting k grids which are uniformly distributed from the n grids to make a uniform design table; establishing a self-organizing feature mapping neural network, and performing normalization processing on data in the uniform design table to serve as the input of the network; connecting weights of neurons of an input layer and an output layer of the self-organizing feature mapping neural network, determining a weight adjusting domain, and updating the weights; assigning an initial value to the learning rate of the self-organizing feature mapping neural network, and setting the learning rate to gradually decrease along with the increase of the training times; and comparing the similarity of the input quantity and the weight corresponding to the neuron in the competition layer by adopting an Euclidean distance method to obtain the sparse position of the sound source, and finally realizing the accurate positioning of the sound source.
Description
Technical Field
The invention relates to the technical field of sound source positioning, in particular to a sound source positioning method based on uniform design and self-organizing feature mapping neural network.
Background
The sound signal in the environment is an important data source for our information acquisition. In space, a sound source is required to be positioned frequently, the position of the sound source in the space is obtained after calculation processing according to collected sound signals, visualization of the sound source is achieved, but many problems cannot be effectively solved when complex sound source environments need to be distinguished and analyzed, and the sound source cannot be positioned quickly and accurately. When an acoustic signal with certain characteristics is transmitted in space, the traditional analysis algorithm for sound source positioning has the disadvantages of large calculation amount, low precision and long positioning time. There is a need for a fast, accurate and adaptive method for determining the intrinsic regularity and nature of signals to locate the position of a sound source in space.
Disclosure of Invention
The technical problem to be solved by the present invention is to provide a sound source localization method based on uniform design and self-organizing feature mapping neural network, so as to realize localization of the overlapped and mixed sound signal source in the three-dimensional space and improve the spatial resolution.
In order to solve the technical problems, the technical scheme adopted by the invention is as follows: the sound source positioning method based on the uniform design and the self-organizing feature mapping neural network comprises the following steps:
step 1, determining a three-dimensional space where a sound source is located, dividing the three-dimensional space where the sound source is located into n three-dimensional grids with the same size according to the overall shape of the three-dimensional space, wherein the body center of each three-dimensional grid has an independent coordinate in a space coordinate system, and the body center coordinate of each three-dimensional grid is used as the position of the sound source to be positioned; the size of the meshing determines the required sound source positioning accuracy;
step 2, determining experimental factors in a uniform design method; selecting m characteristics of the sound signal as experimental factors in a uniform design method; for the sound signals collected in each stereo grid in the step 1, a one-dimensional vector X formed by m experimental factors existsm＝(x1,x2，…xi…xm) However, vectors in different stereoscopic grids, namely at different positions, are different, and for the whole three-dimensional space, n m-dimensional vectors exist;
step 3, selecting k grids which are uniformly distributed from n grids in the three-dimensional space by adopting a uniform design method;
step 4, making the selected k grids into a uniform design table with k rows and m columns, wherein the k rows represent the selected k grids for the positioning experiment, and the m columns are respectively different characteristics of sound signals in the k grids;
step 5, establishing a self-organizing feature mapping neural network; the self-organizing feature mapping neural network is of a two-layer neural network model structure, one layer is an input layer, the other layer is a competition layer, each neuron of the two-layer network is connected in a bidirectional mode, the competition layer is used as an output layer, and each neuron of the two-layer network is connected in a transverse mode; the number of neurons in the input layer is m, the number of neurons corresponds to the number of m experimental factors in the step 2, namely the number of the neurons is equal to the dimension of the sample, and the number of the neurons in the competition layer, namely the number of the neurons in the output layer is k;
step 6, normalizing the vector data of k rows and m columns in the uniform design table
The processed vector is represented as
As an input vector of the self-organizing feature mapping neural network;
step 7, mapping neuron connection weights omega of the input layer and the output layer of the neural network on the self-organizing featuresjAssigning; the connection weight is taken from any k of n m-dimensional vectors in the space, and the connection weight is normalized to obtain a weight vector
Step 8, calculating the product of the input vector and the weight vector of the self-organizing feature mapping neural network to obtain z output values
For z output values computed from the input vector, a win neighborhood N is definedj(t), determining weight adjustment domain for training t times until the winning neighborhood is reduced to have a maximum output value y along with the increase of training timespP ∈ (1, z), called winning neuron; and updating the weights of the winning neurons and other neurons in the weight adjustment domain, wherein the weight adjustment domain comprises the following formula:
wherein, Δ ωjkUpdating variable quantity for the weight of the neuron in the weight adjustment domain, wherein eta is the learning rate of the self-organizing feature mapping neural network;
step 9, assigning an initial value to the learning rate eta of the self-organizing feature mapping neural network, and setting the learning rate to gradually decrease along with the increase of the training times so as to ensure convergence, wherein the formula is as follows:
wherein a is a constant between 0 and 1, t is the training times, and t is the number of training timeslMaximum training times set for the self-organizing feature mapping neural network;
step 10, positioning a sound source; for the input vector, similarity comparison is carried out on the input vector and a weight corresponding to a neuron in a competition layer by adopting an Euclidean distance method; of the k input vectors, the k vectors are,
and the input vector corresponding to the minimum Euclidean distance calculation result is closest to the sound source position, namely the stereo grid where the input vector is located is closest to the sound source position, and then the sound source sparse position is obtained.
Adopt the produced beneficial effect of above-mentioned technical scheme to lie in: the sound source positioning method based on the uniform design and the self-organizing feature mapping neural network, which is provided by the invention, adopts the self-organizing feature mapping neural network to realize sound source positioning, is a calculation method different from the traditional analytic method for sound source positioning, has the advantages of high precision, small operand and accurate positioning, and provides a new idea for space sound source positioning.
Drawings
FIG. 1 is a flow chart of a sound source localization method based on a neural network of uniform design and self-organizing feature mapping according to an embodiment of the present invention;
fig. 2 is a schematic structural diagram of a self-organizing feature mapping neural network according to an embodiment of the present invention.
Detailed Description
The following detailed description of embodiments of the present invention is provided in connection with the accompanying drawings and examples. The following examples are intended to illustrate the invention but are not intended to limit the scope of the invention.
In this embodiment, the sound source localization method based on the uniform design and the self-organizing feature mapping neural network, as shown in fig. 1, includes the following steps:
step 1, determining a three-dimensional space where a sound source is located, dividing the three-dimensional space where the sound source is located into n three-dimensional grids with the same size according to the overall shape of the three-dimensional space, wherein the body center of each three-dimensional grid has an independent coordinate in a space coordinate system, and the body center coordinate of each three-dimensional grid is used as the position of the sound source to be located; the size of the meshing determines the required sound source positioning accuracy; in this embodiment, the three-dimensional space where the sound source is located is a rectangular solid space with length, width and height a, b and c, respectively, and the number of meshes in each direction is la,lb,lcWhen n is equal to la·lb·lc；
Step 2, determining experimental factors in a uniform design method; m characteristics of the sound signals are selected as experimental factors in a uniform design method, and for the sound signals collected in each three-dimensional grid in the step 1, one-dimensional vectors X formed by the m experimental factors existm＝(x1,x2，…xi…xm) However, vectors in different stereoscopic grids, namely at different positions, are different, and for the whole three-dimensional space, n m-dimensional vectors exist; in this embodiment, five characteristics of the frequency, amplitude, phase, sound pressure, and sound intensity of the sound signal are selected as experimental factors in the uniform design method.
Step 3, for the larger three-dimensional space where the sound source is located, the more the number of the divided grids, therefore, a uniform design method is adopted to select k grids which are uniformly distributed from n grids in the three-dimensional space;
step 4, making the selected k grids into a uniform design table with k rows and m columns, wherein the k rows represent the selected k grids for the positioning experiment, and the m columns are respectively different characteristics of sound signals in the k grids;
step 5, establishing a self-organizing feature mapping neural network; the self-organizing feature mapping neural network is of a two-layer neural network model structure, one layer is an input layer, the other layer is a competition layer, each neuron of the two-layer network is connected in a bidirectional mode, the competition layer is used as an output layer, and each neuron of the two-layer network is connected in a transverse mode; the number of neurons in the input layer is m, the number of neurons corresponds to the number of m experimental factors in the step 2, namely the number of the neurons is equal to the dimension of the sample, and the number of the neurons in the competition layer, namely the number of the neurons in the output layer is k;
the Self-Organizing Feature mapping network (SOFM), also called Kohonen network, has two layers, as shown in fig. 2, and is composed of an input layer and an output layer (competition layer), the topology structure of the SOFM does not include a hidden layer, and each neuron of the input layer collects external information to each neuron of the output layer through a weight vector. The number of neurons in the input layer is equal to the sample dimension. The input layer is one-dimensional; the competing layers form a two-dimensional array, which may be one-dimensional, two-dimensional, or multi-dimensional. Wherein the two-dimensional competition layer is constructed in a matrix mode, and the application of the two-dimensional competition layer is the most extensive. All the neurons of the input layer and the competitive layer are connected with each other, and the competitive layers are connected with each other in a lateral direction. The network automatically classifies the input modes according to the learning rules, namely under the condition of no instructor, the network extracts the characteristics of each input mode by self-organizing and learning the input modes and expresses the classification result in a competition layer.
Step 6, normalizing the vector data of k rows and m columns in the uniform design table
The processed vector is represented as
As an input vector of the self-organizing feature mapping neural network;
step 7, mapping neuron connection weights omega of the input layer and the output layer of the neural network on the self-organizing featuresjAssigning; the connection weight is taken from any k of n m-dimensional vectors in the space, and the connection weight is normalized to obtain a weight vector
Step 8, calculating the product of the input vector and the weight vector of the self-organizing feature mapping neural network to obtain each output value z
For input vector calculationThe obtained z output values define a win neighborhood Nj(t), determining weight adjustment domain for training t times until the winning neighborhood is reduced to have a maximum output value y along with the increase of training timespP ∈ (1, z), called winning neuron; and updating the weights of the winning neurons and other neurons in the weight adjustment domain, wherein the weight adjustment domain comprises the following formula:
wherein, Δ ωjkUpdating the variable quantity for the weight in the weight adjustment domain, wherein eta is the learning rate of the self-organizing feature mapping neural network;
step 9, assigning an initial value to the learning rate eta of the self-organizing feature mapping neural network, and setting the learning rate to gradually decrease along with the increase of the training times so as to ensure convergence, wherein the formula is as follows:
wherein a is a constant between 0 and 1, t is the training times, and t is the number of training timeslMaximum training times set for the self-organizing feature mapping neural network;
as can be seen from the calculation formula of the learning rate eta, the learning rate gradually decreases with the increase of the training times, and the fine adjustment of the output is realized.
Step 10, positioning a sound source; for the input vector, similarity comparison is carried out on the input vector and a weight corresponding to a neuron in a competition layer by adopting an Euclidean distance method; of the k input vectors, the k vectors are,
and the input vector corresponding to the minimum Euclidean distance calculation result is closest to the sound source position, namely the closer the stereo grid where the input vector is located is to the sound source position, so as to obtain the sound source sparse position.
Finally, it should be noted that: the above examples are only intended to illustrate the technical solution of the present invention, but not to limit it; although the present invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some or all of the technical features may be equivalently replaced; such modifications and substitutions do not depart from the spirit of the corresponding technical solutions and scope of the present invention as defined in the appended claims.
Claims (6)
1. A sound source positioning method based on a uniform design and a self-organizing feature mapping neural network is characterized in that: the method comprises the following steps:
step 1, determining a three-dimensional space where a sound source is located, dividing the three-dimensional space where the sound source is located into n three-dimensional grids with the same size according to the overall shape of the three-dimensional space, wherein the body center of each three-dimensional grid has an independent coordinate in a space coordinate system, and the body center coordinate of each three-dimensional grid is used as the position of the sound source to be positioned;
step 2, determining experimental factors in a uniform design method; selecting m characteristics of the sound signal as experimental factors in a uniform design method; for the sound signals collected in each stereo grid in the step 1, a one-dimensional vector X formed by m experimental factors existsm＝(x1,x2，…xi…xm) However, vectors in different stereoscopic grids, namely at different positions, are different, and for the whole three-dimensional space, n m-dimensional vectors exist;
step 3, selecting k grids which are uniformly distributed from n grids in the three-dimensional space by adopting a uniform design method;
step 4, making the selected k grids into a uniform design table with k rows and m columns, wherein the k rows represent the selected k grids for the positioning experiment, and the m columns are respectively different characteristics of sound signals in the k grids;
step 5, establishing a self-organizing feature mapping neural network;
step 6, normalizing the vector data of k rows and m columns in the uniform design table, and expressing the processed vector as
As an input vector of the self-organizing feature mapping neural network;
step 7, mapping neuron connection weights omega of the input layer and the output layer of the neural network on the self-organizing featuresjAssigning; the connection weight is taken from any k of n m-dimensional vectors in the space, and the connection weight is normalized to obtain a weight vector
Step 8, calculating the product of the input vector and the weight vector of the self-organizing feature mapping neural network to obtain z output values
For z output values computed from the input vector, a win neighborhood N is definedj(t), determining weight adjustment domain for training t times until the winning neighborhood is reduced to have a maximum output value y along with the increase of training timespP ∈ (1, z), called winning neuron; updating the weights of the winning neurons and other neurons in the weight adjustment domain;
step 9, assigning an initial value to the learning rate eta of the self-organizing feature mapping neural network, and setting the learning rate to gradually decrease along with the increase of the training times so as to ensure convergence;
step 10, positioning a sound source; for the input vector, similarity comparison is carried out on the input vector and a weight corresponding to a neuron in a competition layer by adopting an Euclidean distance method; in the k input vectors, the input vector corresponding to the minimum Euclidean distance calculation result is closest to the sound source position, namely the three-dimensional grid where the input vector is located is closest to the sound source position, and then the sound source sparse position is obtained.
2. The sound source localization method based on uniform design and self-organizing feature mapping neural network of claim 1, wherein: the size of the meshing determines the required sound source localization accuracy.
3. The sound source localization method based on uniform design and self-organizing feature mapping neural network of claim 1, wherein: step 5, the self-organizing feature mapping neural network is of a two-layer neural network model structure, wherein one layer is an input layer, the other layer is a competition layer, the neurons of the two layers of networks are connected in a two-way mode, the competition layer is used as an output layer, and the neurons of the layer are connected in a transverse mode; the number of neurons in the input layer is m, the number of neurons corresponds to the number of m experimental factors in the step 2, namely the number of the neurons is equal to the dimension of the sample, and the number of the neurons in the competition layer, namely the number of the neurons in the output layer is k.
4. The sound source localization method based on uniform design and self-organizing feature mapping neural network of claim 3, wherein: step 8, updating the weights of the winning neurons and other neurons in the weight adjustment domain, which is shown in the following formula:
wherein, Δ ωjkAnd (3) updating the variable quantity of the weight of the neuron in the weight adjustment domain, wherein eta is the learning rate of the self-organizing feature mapping neural network.
5. The sound source localization method based on uniform design and self-organizing feature mapping neural network of claim 4, wherein: the learning rate η of the self-organizing feature mapping neural network in the step 9 is shown as the following formula:
wherein a is a constant between 0 and 1, t is the training times, and t is the number of training timeslAnd setting the maximum training times for the self-organizing feature mapping neural network.
6. The sound source localization method based on uniform design and self-organizing feature mapping neural network of claim 5, wherein: for the input vector, the Euclidean distance method is adopted to compare the similarity of the input vector and the weight corresponding to the neuron in the competition layer, and the following formula is shown:
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010021009.6A CN111239685B (en) | 2020-01-09 | 2020-01-09 | Sound source positioning method based on uniform design and self-organizing feature mapping neural network |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010021009.6A CN111239685B (en) | 2020-01-09 | 2020-01-09 | Sound source positioning method based on uniform design and self-organizing feature mapping neural network |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111239685A CN111239685A (en) | 2020-06-05 |
| CN111239685B true CN111239685B (en) | 2021-08-24 |
Family
ID=70872516
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010021009.6A Active CN111239685B (en) | 2020-01-09 | 2020-01-09 | Sound source positioning method based on uniform design and self-organizing feature mapping neural network |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111239685B (en) |
Families Citing this family (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111954154B (en) * | 2020-08-19 | 2022-08-12 | 北京京东振世信息技术有限公司 | Positioning method and device, computer readable storage medium and electronic device |
| CN112816884B (en) * | 2021-03-01 | 2022-08-02 | 中国人民解放军国防科技大学 | Method, device and equipment for monitoring health state of satellite lithium ion battery |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2000073815A1 (en) * | 1999-05-28 | 2000-12-07 | Fisher-Tortolano Systems Limited | Navigation apparatus and method |
| CN103077267A (en) * | 2012-12-28 | 2013-05-01 | 电子科技大学 | Parameter sound source modeling method based on improved BP (Back Propagation) neural network |
| CN103837859A (en) * | 2014-03-05 | 2014-06-04 | 湖州师范学院 | Method for calculating position of sound emission source through wavelet neural network of mixed leapfrog |
| CN106405492A (en) * | 2016-08-31 | 2017-02-15 | 广西科技大学 | Mobile target detection method based on acoustic waves and wireless positioning |
| CN106940439A (en) * | 2017-03-01 | 2017-07-11 | 西安电子科技大学 | K mean cluster weighting sound localization method based on wireless acoustic sensor network |
| CN108170147A (en) * | 2017-12-31 | 2018-06-15 | 南京邮电大学 | A kind of unmanned plane mission planning method based on self organizing neural network |
| CN109459669A (en) * | 2019-01-09 | 2019-03-12 | 国网上海市电力公司 | 10kV one-phase earthing failure in electric distribution network Section Location |
| WO2019145923A2 (en) * | 2018-01-29 | 2019-08-01 | Wilfred Edwin Booij | Ad hoc positioning of mobile devices using near ultrasound signals |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| FR2925196B1 (en) * | 2007-12-12 | 2011-05-06 | Univ Paris 13 | VALUES TOPOLOGICAL LINK SELF-ORGANIZING CARDS, METHOD FOR CONTROLLING AN AUTOMATIC LEARNING SYSTEM USING SUCH SELF-ORGANIZING CARD AND DYNAMIC BEHAVIOR ANALYZER. |
| US10402995B2 (en) * | 2017-07-27 | 2019-09-03 | Here Global B.V. | Method, apparatus, and system for real-time object detection using a cursor recurrent neural network |
-
2020
- 2020-01-09 CN CN202010021009.6A patent/CN111239685B/en active Active
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2000073815A1 (en) * | 1999-05-28 | 2000-12-07 | Fisher-Tortolano Systems Limited | Navigation apparatus and method |
| CN103077267A (en) * | 2012-12-28 | 2013-05-01 | 电子科技大学 | Parameter sound source modeling method based on improved BP (Back Propagation) neural network |
| CN103837859A (en) * | 2014-03-05 | 2014-06-04 | 湖州师范学院 | Method for calculating position of sound emission source through wavelet neural network of mixed leapfrog |
| CN106405492A (en) * | 2016-08-31 | 2017-02-15 | 广西科技大学 | Mobile target detection method based on acoustic waves and wireless positioning |
| CN106940439A (en) * | 2017-03-01 | 2017-07-11 | 西安电子科技大学 | K mean cluster weighting sound localization method based on wireless acoustic sensor network |
| CN108170147A (en) * | 2017-12-31 | 2018-06-15 | 南京邮电大学 | A kind of unmanned plane mission planning method based on self organizing neural network |
| WO2019145923A2 (en) * | 2018-01-29 | 2019-08-01 | Wilfred Edwin Booij | Ad hoc positioning of mobile devices using near ultrasound signals |
| CN109459669A (en) * | 2019-01-09 | 2019-03-12 | 国网上海市电力公司 | 10kV one-phase earthing failure in electric distribution network Section Location |
Non-Patent Citations (5)
| Title |
|---|
| 《3-D Sound Source Detection by Binaural Model Using Self Organizing Map》;Zhong Zhang; I. Kazuaki; Tetsuo Miyake; Takashi Imamura;《 2008 3rd International Conference on Innovative Computing Information and Control》;20080822;1-4 * |
| 《Spike source localization with tetrodes》;Mircea I. Chelaru1;《Journal of Neuroscience Methods》;20050330;305-315 * |
| 《一种改进的无线传感器网络DV_Hop定位算法》;沈学利; 陈光;《计算机工程》;20150715;115-119 * |
| 《人工神经网络在车船牌照定位中的应用》;戴慰 等;《中国水运》;20061231;140-141 * |
| 《基于自组织特征映射的多站定位技术研究》;陈慧贤；钟子发；王伦文;《通信对抗》;20050331;1 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN111239685A (en) | 2020-06-05 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110728224B (en) | Remote sensing image classification method based on attention mechanism depth Contourlet network | |
| CN108709745B (en) | Rapid bearing fault identification method based on enhanced LPP algorithm and extreme learning machine | |
| Zhang et al. | Local density adaptive similarity measurement for spectral clustering | |
| CN110751121B (en) | Unsupervised radar signal sorting method based on clustering and SOFM | |
| Yang et al. | Recomputation of the dense layers for performance improvement of dcnn | |
| CN114488140B (en) | Small sample radar one-dimensional image target recognition method based on deep migration learning | |
| CN111239685B (en) | Sound source positioning method based on uniform design and self-organizing feature mapping neural network | |
| CN111415379B (en) | Three-dimensional point cloud data registration method based on cuckoo optimization | |
| CN110321451B (en) | Image retrieval algorithm based on distribution entropy gain loss function | |
| Yang et al. | Underwater acoustic research trends with machine learning: general background | |
| Sabrol et al. | Fuzzy and neural network based tomato plant disease classification using natural outdoor images | |
| CN109063757A (en) | It is diagonally indicated based on block and the multifarious multiple view Subspace clustering method of view | |
| CN114330695B (en) | Phased array driving method based on multilayer perception neural network | |
| JP2019049414A (en) | Sound processing device, sound processing method and program | |
| Tian et al. | Joint learning model for underwater acoustic target recognition | |
| CN110690930A (en) | Information source number detection method and device | |
| CN110503148B (en) | Point cloud object identification method with scale invariance | |
| Mishra et al. | Kohonen self organizing map with modified k-means clustering for high dimensional data set | |
| Qin et al. | Two-stage feature selection for classification of gene expression data based on an improved Salp Swarm Algorithm | |
| CN108549936A (en) | The Enhancement Method that self organizing neural network topology based on deep learning is kept | |
| Ren et al. | Pulses classification based on sparse auto-encoders neural networks | |
| CN113722951B (en) | Scatterer three-dimensional finite element grid optimization method based on neural network | |
| CN113850304A (en) | High-accuracy point cloud data classification segmentation improvement algorithm | |
| CN107563399A (en) | The characteristic weighing Spectral Clustering and system of a kind of knowledge based entropy | |
| Polani et al. | Organization measures for self-organizing maps |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |