CN112270406A - Neural information visualization method of brain-like computer operating system - Google Patents
Neural information visualization method of brain-like computer operating system Download PDFInfo
- Publication number
- CN112270406A CN112270406A CN202011255463.4A CN202011255463A CN112270406A CN 112270406 A CN112270406 A CN 112270406A CN 202011255463 A CN202011255463 A CN 202011255463A CN 112270406 A CN112270406 A CN 112270406A
- Authority
- CN
- China
- Prior art keywords
- neuron
- nodes
- importance
- node
- brain
- 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.)
- Granted
Links
- 230000001537 neural effect Effects 0.000 title claims abstract description 33
- 238000007794 visualization technique Methods 0.000 title claims abstract description 12
- 210000002569 neuron Anatomy 0.000 claims abstract description 147
- 238000013528 artificial neural network Methods 0.000 claims abstract description 44
- 238000012800 visualization Methods 0.000 claims abstract description 24
- 238000000034 method Methods 0.000 claims abstract description 18
- 239000003086 colorant Substances 0.000 claims abstract description 7
- 238000012421 spiking Methods 0.000 claims description 9
- 230000000007 visual effect Effects 0.000 claims description 9
- 230000008569 process Effects 0.000 abstract description 4
- 238000010586 diagram Methods 0.000 description 7
- 238000013527 convolutional neural network Methods 0.000 description 5
- 238000005070 sampling Methods 0.000 description 5
- 238000013135 deep learning Methods 0.000 description 3
- 230000002776 aggregation Effects 0.000 description 2
- 238000004220 aggregation Methods 0.000 description 2
- 238000003062 neural network model Methods 0.000 description 2
- 238000011160 research Methods 0.000 description 2
- 238000007792 addition Methods 0.000 description 1
- 230000006399 behavior Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 210000004556 brain Anatomy 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 239000011159 matrix material Substances 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 238000003672 processing method Methods 0.000 description 1
- 238000004088 simulation Methods 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
Images
Classifications
-
- 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
- G06N3/06—Physical realisation, i.e. hardware implementation of neural networks, neurons or parts of neurons
- G06N3/061—Physical realisation, i.e. hardware implementation of neural networks, neurons or parts of neurons using biological neurons, e.g. biological neurons connected to an integrated circuit
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F9/00—Arrangements for program control, e.g. control units
- G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
- G06F9/44—Arrangements for executing specific programs
- G06F9/451—Execution arrangements for user interfaces
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
- Y02D10/00—Energy efficient computing, e.g. low power processors, power management or thermal management
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Software Systems (AREA)
- Theoretical Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- Biophysics (AREA)
- Biomedical Technology (AREA)
- Health & Medical Sciences (AREA)
- General Physics & Mathematics (AREA)
- Molecular Biology (AREA)
- General Engineering & Computer Science (AREA)
- Neurology (AREA)
- Computational Linguistics (AREA)
- Microelectronics & Electronic Packaging (AREA)
- Artificial Intelligence (AREA)
- Human Computer Interaction (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- General Health & Medical Sciences (AREA)
- Computing Systems (AREA)
- Mathematical Physics (AREA)
- User Interface Of Digital Computer (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses a neural information visualization method of a brain-like computer operating system, which comprises the following steps: obtaining a connection description file of the pulse neural network, and carrying out global coding on each layer of neuron nodes according to the connection description file, wherein each neuron node obtains a global unique identifier, each layer of neuron nodes are represented in a gathering way and are distinguished by different colors, and the connection relation of the neuron nodes between adjacent levels is represented by connecting edges and is distinguished by different colors; and calculating the importance of each neuron node according to the number of the connecting edges of the neuron nodes and the weight of the connecting edges, and selecting the neuron nodes for visualization according to the importance. The problems that a plurality of connection description files are needed, the number of neuron nodes is large, and connection lines among the neuron nodes are large and difficult to distinguish are solved. The neural information of the brain-like computer operating system is reasonably visualized, and the understanding of the working process of the impulse neural network is facilitated.
Description
Technical Field
The invention belongs to the technical field of novel computers, and particularly relates to a neural information visualization method of a brain-like computer operating system.
Background
Since 2006, deep learning has been highly concerned by scientific research institutions and the industry, and has been greatly developed in a plurality of fields such as images and voices, and comprehensively surpasses traditional algorithms in a plurality of fields. Spiking neural Networks (SNN-Spiking neural Networks) are often known as third generation artificial neural Networks. The first generation of neural networks were perceptrons, which were a simple neuron model and could only process binary data. The second generation neural networks include a wide variety of BP neural networks including more applications. And the impulse neural network more closely simulates the connection relation and the behavior of brain neurons.
At present, the tools for visualizing the convolutional neural network structure disclosed on the network are more, and mainly include: 1) netscope: the model definition file is used as input to obtain a visual structure diagram of the neural network, the visual structure diagram is a visual neural network topology tool based on a webpage, and only supports a caffe deep learning framework of Berkeley division of California university. 2) ConvNetDraw: the visualization of the output neural network model structure using visualization commands as input is also a web-based tool. The model of the structural block is used for displaying, the three-dimensional proportion adjustment can be carried out on the structural block, and the method is very visual and visual. 3) Netron: a visual structure diagram of a neural network is obtained by taking a model definition file and a model weight file (which can be default) as input, and the visual structure diagram is also based on a webpage and uses js and python and can support mainstream deep learning frameworks such as ONNX, Keras, CoreML, TensorFlow, caffe and MXNET.
The patent application of Cao Li Macro in China media university with publication number CN106372721A discloses a method for visualizing a large-scale neural network in a 3D form, and shows the structure of the neural network in a 3D form; qiufuyang et al, published under CN107392085A, of Shanghai precision metrology testing institute, discloses a method for visualizing a convolutional neural network, which can well demonstrate the superiority of the convolutional neural network for exploring the convolutional neural network; the patent application with publication number CN110782031A of Beijing computer technology and application research discloses a multi-frame convolutional neural network model structure visualization and network reconstruction method, which can visually modify different layers to realize network reconstruction and can change the attribute of the neural network and update and display the attribute in real time.
Due to the complexity of the impulse neural network structure of the brain-like computer operating system, when the impulse neural network structure of the brain-like computer operating system is visualized, the problems of more connection description files, more neuron nodes and more connecting lines among the neuron nodes are difficult to distinguish. Therefore, all the visualization methods described above are not suitable for visualization of the spiking neural network structure of the brain-like computer operating system.
Disclosure of Invention
The invention aims to provide a neural information visualization method of a brain-like computer operating system, which aims to solve the problems that a plurality of connection description files are provided, the number of neuron nodes is large, and a plurality of connecting lines among the neuron nodes are difficult to distinguish.
In order to achieve the purpose, the invention provides the following technical scheme:
a method for visualizing neural information of a brain-like computer operating system, comprising the steps of:
obtaining a connection description file of the pulse neural network, and carrying out global coding on each layer of neuron nodes according to the connection description file, wherein each neuron node obtains a global unique identifier, each layer of neuron nodes are represented in a gathering way and are distinguished by different colors, and the connection relation of the neuron nodes between adjacent levels is represented by connecting edges and is distinguished by different colors;
and calculating the importance of each neuron node according to the number of the connecting edges of the neuron nodes and the weight of the connecting edges, and selecting the neuron nodes for visualization according to the importance.
Preferably, all the neuron nodes in the same layer of the spiking neural network are represented in an aggregation mode, and the neuron nodes in two adjacent layers are represented in the same color, and the aggregation representations of the neuron nodes in two adjacent layers are also located in adjacent positions, so that the connection relation between the neuron nodes in two adjacent layers can be represented conveniently.
Preferably, each layer of neuron nodes is aggregated into a geometric shape for representation. The geometric shape can be a smooth geometric shape without edges and corners, such as a circle, an ellipse and the like, and can also be a geometric shape with edges and corners, such as a matrix, a triangle and the like.
Preferably, all connection relationships between two layers of neuron nodes are represented by connecting edges in the same color. In order to improve the visualization definition, the color of the neuron node is different from the color of the corresponding connecting edge.
Preferably, when the neural information visualization is performed, the neuron information layout is stored in a data structure, and the data structure comprises a data unit, a connection relation unit and a layer unit, wherein the data unit is used for storing the global unique identifier, the visualization position coordinate and the level of the neuron node; the connection relation unit is used for storing two neuron nodes and the visual color of a connection edge between the two neuron nodes; the layer unit is used for storing the hierarchy name.
Because the number of the neuron nodes is large, whether a certain neuron is displayed or not is selected according to the number of the connecting edges of the neuron nodes and other neurons and the weight of the connecting edges. If too much sampling is carried out on the neuron nodes, the neuron nodes are too dense after visualization, and the connection relation among the neuron nodes cannot be seen; if the sampling of the neuron nodes is too little, the loss of main information can be caused, and the main structure of the impulse neural network cannot be observed. In order to perform reasonable sparse sampling on neurons, the importance of the neuron nodes needs to be calculated and ranked. Preferably, the importance of each neuron node is calculated using the following formula:
wherein ,representing the ith neuron nodeiThe absolute importance of the preceding neuron, InNum represents the in-degree of the neuron node, w1Is the in degree coefficient of the neuron node, OutNum represents the out degree of the neuron node, w2Is the out degree coefficient of the neuron node, Weight represents the sum of absolute values of weights of the connected edges of the neuron node, w3Is the weight coefficient of the neuron node,is the importance of the final neuron node,represents the jth neuron nodejThe absolute significance of the previous neuron, n represents the total number of neuron nodes. The number of the connecting edges accessed to the current neuron node is used as the in degree of the current neuron node, and the number of the connecting edges accessed from the current neuron node is used as the out degree of the current neuron node.
Preferably, when the neuron nodes are selected according to the importance degree for visualization, an importance threshold value is set, and the neuron nodes with the importance greater than the importance threshold value are screened for visualization. Thus, the more edges the neuron node is connected to, the larger the absolute value of the weight of the edge is, the higher the weight is, the priority is kept.
Compared with the prior art, the invention has the beneficial effects that at least:
according to the neural information visualization method of the brain-like computer operating system, the problem of a large number of connection description files is solved by carrying out global coding on neurons, the neurons are screened and visualized according to the importance of the neuron nodes, the problem of difficulty in displaying a large number of neuron nodes is solved by adopting the sparseness processing method, meanwhile, the connection relation is expressed in a connection form and is distinguished by different colors, and the problem of difficulty in distinguishing a large number of connection lines among the neuron nodes is solved. The neural information of the brain-like computer operating system is reasonably visualized, and the understanding of the working process of the impulse neural network is facilitated.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, it is obvious that the drawings in the following description are only some embodiments of the present invention, and for those skilled in the art, other drawings can be obtained according to these drawings without creative efforts.
FIG. 1 is a flowchart of a method for visualizing neural information of a brain-like computer operating system according to an embodiment of the present invention;
FIG. 2 is an internal data structure of a connection description file of the spiking neural network provided by an embodiment of the present invention;
FIG. 3 is an internal data structure of an intermediate json file for visualization of neural information provided by an embodiment of the present invention;
FIG. 4 is a diagram of a quadruple within a data field of a json file according to an embodiment of the present invention;
FIG. 5 is a triple inside links fields of a json file provided by an embodiment of the present invention;
FIG. 6 is a neural information visualization diagram of an electroencephalogram simulated impulse neural network provided by an embodiment of the invention;
fig. 7 is a neural information visualization diagram of a memory model impulse neural network provided by an embodiment of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention will be further described in detail with reference to the accompanying drawings and examples. It should be understood that the detailed description and specific examples, while indicating the scope of the invention, are intended for purposes of illustration only and are not intended to limit the scope of the invention.
In the process of carrying out neural information visualization on the impulse neural network of the brain-like computer operating system of Darwin No. 2 neural mimicry chip in Zhejiang university, the problems of more connection description files, more neuron nodes and more connecting lines among the neuron nodes are encountered, and the connection description files are not easy to distinguish. In order to solve the problem that neural information visualization is difficult due to the problems, the embodiment of the invention provides a neural information visualization method of a brain-like computer operating system, which is used for performing neural information visualization based on a pulse neural network model connection description file of the brain-like computer operating system of a Darwin No. 2 neural mimicry chip.
Fig. 1 is a flowchart of a method for visualizing neural information of a brain-like computer operating system according to an embodiment of the present invention. As shown in fig. 1, the neural information visualization method includes the steps of:
step 1, reading a connection description file of the impulse neural network according to the connection relation between layers of the impulse neural network and the sequence from an input layer to an output layer.
The connection description file of the impulse neural network describes the connection relationship between the neurons of the impulse neural network and the neurons. The connecting description file of the impulse neural network is multiple, and each file describes the connecting relation between two layers of neurons. Fig. 2 is an internal data structure of a connection description file of the impulse neural network. As shown in fig. 2, each row of the connection description file is a quadruple, the first field of the quadruple is the intra-layer number of the starting neuron, i.e. the number of the starting neuron in the starting layer, the second field is the intra-layer number of the target neuron, i.e. the number of the target neuron in the target layer, the third field is the weight, i.e. the weight of the processing input value of the neuron, which is the inherent property of the connecting edge of the neuron, and the fourth field is the pulse delay, i.e. the delay of the pulse signal from the starting neuron to the target neuron.
And 2, according to the read pulse neural network connection description file, globally numbering each layer of neuron nodes again, converting the intra-layer numbers into globally unique identifiers, and simultaneously determining the color of each layer of neurons to form the attributes of the neurons.
And 3, generating a triple (initial neuron number, target neuron number and edge color) for describing the node connection relation according to the connection description file and the self attribute of the neuron node, wherein the edge color is determined according to the layer where the edge connection neuron is located.
And 4, distributing the neuron nodes according to the connection relation between the neurons of the pulse neural network, wherein the neuron nodes of the same layer are gathered and placed in the same area, and the neurons of the adjacent layer are gathered and placed at the adjacent positions.
And 5, selecting whether a certain neuron is displayed or not according to the number of the connecting edges of the neuron and other neurons and the weight of the connecting edges.
Because the number of neuron nodes and connecting lines in the original spiking neural network is too many, if all the neuron nodes and connecting lines are displayed completely, all the node nodes and connecting lines are crowded together, and the key structure of the spiking neural network cannot be displayed, so that the spiking neural network needs to be displayed sparsely. If the sampling of the neurons is too much, the visualized neurons are too dense, and the connection relation among the neurons cannot be seen; if the sampling of the neuron is too little, the loss of main information can be caused, and the main structure of the impulse neural network cannot be observed. The method specifically comprises the steps of calculating the importance of the neurons according to the number of the connecting edges of the neurons and other neurons and the weight of the connecting edges, and sorting the neurons according to the importance, wherein the neurons with the greater importance are preferentially reserved.
The formula for the significance of neurons is as follows:
after calculating the importance of the neurons, setting an importance threshold value E (0,1)) to control which neurons are displayed, wherein the neurons with the importance greater than the threshold value are displayed, and the neurons smaller than the threshold value are hidden.
And 6, exporting a json file for describing the color of the neuron layout and the edge, and visualizing the exported json file by utilizing echarts.
In the embodiment, a json file is generated by adopting a python code, and the json file is visualized by adopting a graph type of echarts. FIG. 3 is a data structure of a json file, which contains three fields: and the data unit data is used for connecting the relationship units links and the layer units layers. The data field is an array, each internal element is a quadruple, and the quadruple contains the key value name, the position coordinate (x, y) and the belonging layer information of the neuron node, as shown in fig. 4. The links field is also an array, each element inside the array is a triplet, and the triplet includes the key value source of the starting neuron node, the key value target of the target neuron, and the color lineStyle of the link, as shown in fig. 5.
The neural information visualization results of the electroencephalogram simulation impulse neural network and the memory model impulse neural network by using the neural information visualization method of the brain-like computer operating system are shown in fig. 6 and 7, wherein the structure of the impulse neural network is clearly shown in the figures, namely, the connection relation between the neuron nodes of the impulse neural network and each neuron node is shown.
The above-mentioned embodiments are intended to illustrate the technical solutions and advantages of the present invention, and it should be understood that the above-mentioned embodiments are only the most preferred embodiments of the present invention, and are not intended to limit the present invention, and any modifications, additions, equivalents, etc. made within the scope of the principles of the present invention should be included in the scope of the present invention.
Claims (8)
1. A neural information visualization method of a brain-like computer operating system, comprising the steps of:
obtaining a connection description file of the pulse neural network, and carrying out global coding on each layer of neuron nodes according to the connection description file, wherein each neuron node obtains a global unique identifier, each layer of neuron nodes are represented in a gathering way and are distinguished by different colors, and the connection relation of the neuron nodes between adjacent levels is represented by connecting edges and is distinguished by different colors;
and calculating the importance of each neuron node according to the number of the connecting edges of the neuron nodes and the weight of the connecting edges, and selecting the neuron nodes for visualization according to the importance.
2. The method for visualizing neural information of a brain-like computer operating system of claim 1, wherein all the neuron nodes in the same layer of the spiking neural network are represented in an aggregate manner and in the same color, and the aggregate representations of the neuron nodes in two adjacent layers are also in adjacent positions, so as to facilitate the representation of the connection relationship between the neuron nodes in two adjacent layers.
3. The method for visualizing neural information of a brain-like computer operating system of claim 1 or 2, wherein each layer of neuron nodes are clustered into geometric shapes for representation.
4. The method for visualizing neural information of a brain-like computer operating system of claim 1, wherein all connections between neuron nodes of two layers are represented by connecting edges in the same color.
5. The method of claim 1, wherein the color of the neuron node is different from the color of the corresponding connecting edge.
6. The neural information visualization method of the brain-like computer operating system according to claim 1, wherein in the neural information visualization, the neural information layout is stored in a data structure, the data structure comprises data units, connection relation units and layer units, wherein the data units are used for storing globally unique identifiers of the neural nodes, visualization position coordinates and the hierarchy; the connection relation unit is used for storing two neuron nodes and the visual color of a connection edge between the two neuron nodes; the layer unit is used for storing the hierarchy name.
7. The method of visualizing neural information of a brain-like computer operating system of claim 1, wherein the importance of each neuron node is calculated using the following formula:
wherein ,representing the ith neuron nodeiThe absolute importance of the preceding neuron, InNum represents the in-degree of the neuron node, w1Is the in degree coefficient of the neuron node, OutNum represents the out degree of the neuron node, w2Is the out degree coefficient of the neuron node, Weight represents the sum of absolute values of weights of the connected edges of the neuron node, w3Is the weight coefficient of the neuron node,is the importance of the final neuron node,represents the jth neuron nodejThe absolute significance of the previous neuron, n represents the total number of neuron nodes.
8. The method as claimed in claim 1, wherein when the neuron nodes are selected for visualization according to their importance, an importance threshold is set, and neuron nodes with importance greater than the importance threshold are selected for visualization.
Priority Applications (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202011255463.4A CN112270406B (en) | 2020-11-11 | 2020-11-11 | Nerve information visualization method of brain-like computer operating system |
PCT/CN2020/128463 WO2022099557A1 (en) | 2020-11-11 | 2020-11-12 | Neuron information visualization method for operating system of brain-like computer |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202011255463.4A CN112270406B (en) | 2020-11-11 | 2020-11-11 | Nerve information visualization method of brain-like computer operating system |
Publications (2)
Publication Number | Publication Date |
---|---|
CN112270406A true CN112270406A (en) | 2021-01-26 |
CN112270406B CN112270406B (en) | 2023-05-23 |
Family
ID=74340656
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202011255463.4A Active CN112270406B (en) | 2020-11-11 | 2020-11-11 | Nerve information visualization method of brain-like computer operating system |
Country Status (2)
Country | Link |
---|---|
CN (1) | CN112270406B (en) |
WO (1) | WO2022099557A1 (en) |
Cited By (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN112561043A (en) * | 2021-03-01 | 2021-03-26 | 浙江大学 | Neural model splitting method of brain-like computer operating system |
CN112561042A (en) * | 2021-03-01 | 2021-03-26 | 浙江大学 | Neural model mapping method of brain-like computer operating system |
CN112819160A (en) * | 2021-02-24 | 2021-05-18 | 文远鄂行(湖北)出行科技有限公司 | Visualization method, device, equipment and storage medium of neural network model |
CN113297314A (en) * | 2021-07-28 | 2021-08-24 | 深圳市永达电子信息股份有限公司 | Data visualization method and device and storage medium |
CN113791913A (en) * | 2021-11-16 | 2021-12-14 | 浙江大学 | Task migration method of brain-like computer operating system |
CN114399033A (en) * | 2022-03-25 | 2022-04-26 | 浙江大学 | Brain-like computing system and method based on neuron instruction coding |
CN114611686A (en) * | 2022-05-12 | 2022-06-10 | 之江实验室 | Synapse delay implementation system and method based on programmable neural mimicry core |
Families Citing this family (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN115879544B (en) * | 2023-02-28 | 2023-06-16 | 中国电子科技南湖研究院 | Neuron coding method and system for distributed brain-like simulation |
Citations (17)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104955388A (en) * | 2012-11-13 | 2015-09-30 | 艾欧敏达有限公司 | Neurophysiological data analysis using spatiotemporal parcellation |
CN105788608A (en) * | 2016-03-03 | 2016-07-20 | 渤海大学 | Chinese initial consonant and compound vowel visualization method based on neural network |
CN106372721A (en) * | 2016-08-29 | 2017-02-01 | 中国传媒大学 | Large-scale nerve network 3D visualization method |
CN106650922A (en) * | 2016-09-29 | 2017-05-10 | 清华大学 | Hardware neural network conversion method, computing device, compiling method and neural network software and hardware collaboration system |
US20170351401A1 (en) * | 2016-06-06 | 2017-12-07 | Salesforce.Com, Inc. | Visualizing Neural Networks |
CN107636693A (en) * | 2015-03-20 | 2018-01-26 | 弗劳恩霍夫应用研究促进协会 | Assigned for the relevance scores of artificial neural network |
CN108140025A (en) * | 2015-05-26 | 2018-06-08 | 阿雅斯迪公司 | For the interpretation of result of graphic hotsopt |
US20190005384A1 (en) * | 2017-06-29 | 2019-01-03 | General Electric Company | Topology aware graph neural nets |
US20190108436A1 (en) * | 2017-10-06 | 2019-04-11 | Deepcube Ltd | System and method for compact and efficient sparse neural networks |
CN110322010A (en) * | 2019-07-02 | 2019-10-11 | 深圳忆海原识科技有限公司 | The impulsive neural networks arithmetic system and method calculated for class brain intelligence with cognition |
US20200026992A1 (en) * | 2016-09-29 | 2020-01-23 | Tsinghua University | Hardware neural network conversion method, computing device, compiling method and neural network software and hardware collaboration system |
CN110781933A (en) * | 2019-10-14 | 2020-02-11 | 杭州电子科技大学 | Visual analysis method for understanding graph convolution neural network |
CN110782031A (en) * | 2019-09-27 | 2020-02-11 | 北京计算机技术及应用研究所 | Multi-frame convolutional neural network model structure visualization and network reconstruction method |
CN110909869A (en) * | 2019-11-21 | 2020-03-24 | 浙江大学 | Brain-like computing chip based on impulse neural network |
CN111082949A (en) * | 2019-10-29 | 2020-04-28 | 广东工业大学 | Method for efficiently transmitting pulse data packets in brain-like computer |
CN111291243A (en) * | 2019-12-30 | 2020-06-16 | 浙江大学 | Visual reasoning method for uncertainty of spatiotemporal information of character event |
CN111797333A (en) * | 2020-06-04 | 2020-10-20 | 南京擎盾信息科技有限公司 | Public opinion spreading task display method and device |
Family Cites Families (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US10055434B2 (en) * | 2013-10-16 | 2018-08-21 | University Of Tennessee Research Foundation | Method and apparatus for providing random selection and long-term potentiation and depression in an artificial network |
CN107704925B (en) * | 2017-10-16 | 2020-04-14 | 清华大学 | Visual analysis system and method for deep neural network training process |
-
2020
- 2020-11-11 CN CN202011255463.4A patent/CN112270406B/en active Active
- 2020-11-12 WO PCT/CN2020/128463 patent/WO2022099557A1/en active Application Filing
Patent Citations (17)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104955388A (en) * | 2012-11-13 | 2015-09-30 | 艾欧敏达有限公司 | Neurophysiological data analysis using spatiotemporal parcellation |
CN107636693A (en) * | 2015-03-20 | 2018-01-26 | 弗劳恩霍夫应用研究促进协会 | Assigned for the relevance scores of artificial neural network |
CN108140025A (en) * | 2015-05-26 | 2018-06-08 | 阿雅斯迪公司 | For the interpretation of result of graphic hotsopt |
CN105788608A (en) * | 2016-03-03 | 2016-07-20 | 渤海大学 | Chinese initial consonant and compound vowel visualization method based on neural network |
US20170351401A1 (en) * | 2016-06-06 | 2017-12-07 | Salesforce.Com, Inc. | Visualizing Neural Networks |
CN106372721A (en) * | 2016-08-29 | 2017-02-01 | 中国传媒大学 | Large-scale nerve network 3D visualization method |
US20200026992A1 (en) * | 2016-09-29 | 2020-01-23 | Tsinghua University | Hardware neural network conversion method, computing device, compiling method and neural network software and hardware collaboration system |
CN106650922A (en) * | 2016-09-29 | 2017-05-10 | 清华大学 | Hardware neural network conversion method, computing device, compiling method and neural network software and hardware collaboration system |
US20190005384A1 (en) * | 2017-06-29 | 2019-01-03 | General Electric Company | Topology aware graph neural nets |
US20190108436A1 (en) * | 2017-10-06 | 2019-04-11 | Deepcube Ltd | System and method for compact and efficient sparse neural networks |
CN110322010A (en) * | 2019-07-02 | 2019-10-11 | 深圳忆海原识科技有限公司 | The impulsive neural networks arithmetic system and method calculated for class brain intelligence with cognition |
CN110782031A (en) * | 2019-09-27 | 2020-02-11 | 北京计算机技术及应用研究所 | Multi-frame convolutional neural network model structure visualization and network reconstruction method |
CN110781933A (en) * | 2019-10-14 | 2020-02-11 | 杭州电子科技大学 | Visual analysis method for understanding graph convolution neural network |
CN111082949A (en) * | 2019-10-29 | 2020-04-28 | 广东工业大学 | Method for efficiently transmitting pulse data packets in brain-like computer |
CN110909869A (en) * | 2019-11-21 | 2020-03-24 | 浙江大学 | Brain-like computing chip based on impulse neural network |
CN111291243A (en) * | 2019-12-30 | 2020-06-16 | 浙江大学 | Visual reasoning method for uncertainty of spatiotemporal information of character event |
CN111797333A (en) * | 2020-06-04 | 2020-10-20 | 南京擎盾信息科技有限公司 | Public opinion spreading task display method and device |
Non-Patent Citations (2)
Title |
---|
RAJA MUBASHAR KARIM等: "A Study of Colormaps in Network Visualization", 《APPLIED SCIENCES》 * |
刘书新: "基于非对称神经网络的跨网络社交关系预测研究", 《中国优秀博硕士学位论文全文数据库(硕士)社会科学Ⅱ辑》 * |
Cited By (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN112819160A (en) * | 2021-02-24 | 2021-05-18 | 文远鄂行(湖北)出行科技有限公司 | Visualization method, device, equipment and storage medium of neural network model |
CN112819160B (en) * | 2021-02-24 | 2023-10-31 | 文远鄂行(湖北)出行科技有限公司 | Visualization method, device and equipment for neural network model and storage medium |
CN112561043A (en) * | 2021-03-01 | 2021-03-26 | 浙江大学 | Neural model splitting method of brain-like computer operating system |
CN112561042A (en) * | 2021-03-01 | 2021-03-26 | 浙江大学 | Neural model mapping method of brain-like computer operating system |
CN112561043B (en) * | 2021-03-01 | 2021-06-29 | 浙江大学 | Neural model splitting method of brain-like computer operating system |
CN113297314A (en) * | 2021-07-28 | 2021-08-24 | 深圳市永达电子信息股份有限公司 | Data visualization method and device and storage medium |
CN113791913A (en) * | 2021-11-16 | 2021-12-14 | 浙江大学 | Task migration method of brain-like computer operating system |
CN114399033A (en) * | 2022-03-25 | 2022-04-26 | 浙江大学 | Brain-like computing system and method based on neuron instruction coding |
CN114611686A (en) * | 2022-05-12 | 2022-06-10 | 之江实验室 | Synapse delay implementation system and method based on programmable neural mimicry core |
Also Published As
Publication number | Publication date |
---|---|
WO2022099557A1 (en) | 2022-05-19 |
CN112270406B (en) | 2023-05-23 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN112270406A (en) | Neural information visualization method of brain-like computer operating system | |
Gyulassy et al. | Efficient computation of Morse-Smale complexes for three-dimensional scalar functions | |
US8605092B2 (en) | Method and apparatus of animation planning for a dynamic graph | |
Van Ham et al. | Interactive visualization of state transition systems | |
CN106778894A (en) | A kind of method of author's cooperative relationship prediction in academic Heterogeneous Information network | |
CN104616207A (en) | Power grid topological visualization system and power grid topological visualization method | |
CN106022469A (en) | Method for scalable neurosynaptic networks | |
CN114169231A (en) | Method for obtaining distribution line fault classification, positioning and line selection deep learning model based on transfer learning | |
CN106226817B (en) | More attribute earthquake data visualization methods based on continuous parallel coordinates | |
Cheng et al. | TorusVis^ ND: unraveling high-dimensional torus networks for network traffic visualizations | |
CN104658033A (en) | Method and device for global illumination rendering under multiple light sources | |
CN115757604B (en) | GDP space-time evolution analysis method based on noctilucent image data | |
CN114492770A (en) | Brain-like computing chip mapping method for cyclic pulse neural network | |
CN112448829A (en) | Network topology display method and device, readable storage medium and intelligent terminal | |
CN109344439A (en) | A kind of modeling of building block formula and its simulative construction method based on BIM | |
CN110705650A (en) | Metal plate layout method based on deep learning | |
CN107480364B (en) | BIM (building information modeling) model dynamic organization method for engineering construction information management | |
CN114254467B (en) | Method for generating power grid ring network graph | |
CN114092641A (en) | Method, device and equipment for generating wiring board drawing and readable storage medium | |
CN111783832B (en) | Interactive selection method of space-time data prediction model | |
CN116068329A (en) | Transmission line fault classification and positioning method based on transfer learning | |
Miłek et al. | Comparative GIS analysis using taxonomy and classification techniques | |
Giordano et al. | A conceptual model of GIS-based spatial analysis | |
Su et al. | A multi-level visualization method for IT system structure | |
Li et al. | Composite feature and variational design concepts in a feature-based design system |
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 |