CN101894295A - Method for simulating attention mobility by using neural network - Google Patents
Method for simulating attention mobility by using neural network Download PDFInfo
- Publication number
- CN101894295A CN101894295A CN2010101992275A CN201010199227A CN101894295A CN 101894295 A CN101894295 A CN 101894295A CN 2010101992275 A CN2010101992275 A CN 2010101992275A CN 201010199227 A CN201010199227 A CN 201010199227A CN 101894295 A CN101894295 A CN 101894295A
- Authority
- CN
- China
- Prior art keywords
- neuron
- network
- attention
- layer
- oscillator
- 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
Images
Landscapes
- Image Analysis (AREA)
Abstract
The invention discloses a method for simulating attention mobility by using a neural network, which adopts a visual image input layer, neuronal oscillator network oscillation layer and an attention mobility realizing layer, wherein the visual image input layer inputs a gray value of a gray level image into a neurodynamic network; the neuronal oscillator network oscillation layer couples dynamic systems, which are built by all oscillators in the neurodynamic network according to a FitzHugh-Nagumo model, to form the neurodynamic network; and the attention mobility realizing layer realizes the attention mobility between the synchronized different objects by changing parameters, so that the frequency output by nerves of a currently concerned object increases. On the basis of the neurodynamic system, the method forms simple simulation on human-eye attention mobility visual treatment by analyzing and modifying the FitzHugh-Nagumo model and has a great theoretical and practical significance for further research on the visual treatment mechanism of human.
Description
Technical field
The present invention relates to Flame Image Process and nerual network technique based on neurodynamics, particularly relate to the neural power network that a kind of neuron oscillator with the neurodynamics model description forms, be used for simulating the method for simple artificial gray level image people's attention mobility.
Background technology
Vision attention is selected and the vision attention transfer is the important mechanisms that guarantees that biosystem is finished the work with limited processing power.Extract notable attribute from image or receptive field, and they are divided into different zones, therefrom selecting significant zone is the basic task that perception is understood.This ability is exactly the vision attention selection during vision is understood.After significant zone was selected, because the adaptability of vision system, notice can forward next significant zone from current significant zone to, and this is attention mobility.In the Flame Image Process based on the neuroid of synchronized oscillation by more and more widely application, the neuron models of pulse coupling have obtained good achievement, but the application in vision attention selection and vision attention transfer remains very important concerning current research
Human eye can carry out perception and understanding when handling visual pattern in a short period of time easily, and realize from a significant zone-transfer to the next one zone significantly, although this research on the one hand is a lot of, yet the researcher still knows little about it so far to mechanism wherein.Researchers have proposed neuron behavior in the real human brain of a lot of neuron oscillator network modelings, and wherein nineteen fifty-two is first model of extensively being quoted by the Hodgkin-Huxley model (hereinafter to be referred as the HH model) that Hodgkin and Huxley propose on the bases of a large amount of experiments.FitzHugh-Nagumo (hereinafter to be referred as FHN) model is from the HH model simplification, and is proved the current potential granting of imictron preferably by a plurality of researchers, is the important models of anthropomorphic dummy's vision.
The FHN model description neuron by input current I
InputFilm potential V that drives and the relation between the internal state variable R.V and R are respectively the rate of change of time t:
V is the potential difference (PD) of neuronal cell film both sides, is called film potential, and R is an internal state variable of representing the current potential threshold value, I
InputBeing the input current that neuron receives, also is this neuronic light stimulus of outer bound pair.Here 10 and 0.8 is the inverse of the time constant of V and R, and α>0, and β>0 has been described the action intensity from R to V and from V to R respectively.The time constant of V is 12.5 times of R, and this has reacted the activation fact more faster than rejuvenation in the aixs cylinder.Suppose that (V R) is equilibrium point, has only that (V when being unsettled R), produces stable limit cycle around it when equilibrium point.Therefore the value of α and β must satisfy formula (2)
8(V
2+αβ-1)<0(2)
Perhaps formula (3)
8(V
2+αβ-1)>0
-(10V
2-9.2)>0(3)
The parameter alpha and the β that change in the equation (1) can control the granting frequency.The increase of α or β all can cause the increase of providing frequency, but when α or β are a quite little value (for example α=1, β=0.6 or α=0.3, β=2 or α=0.3, β=0.6), equation (1) does not produce granting.
Summary of the invention
The objective of the invention is to, by a kind of method with the neuron network simulation attention mobility is provided, use the FHN model of transforming and form neural power network anthropomorphic dummy eye process from a salient region to another in receptive field, and make up attention mobility system in the simple receptive field, single neuron is carried out model construction, and based on single model construction biological vision neural network.
The present invention is to adopt following technological means to realize with the method for neuron network simulation attention mobility.
Step 1, visual pattern input layer are input to the gray-scale value of gray level image in the neural power network, there is one-to-one relationship in neuron oscillator on pixel in the image and the neural power network, and the neurodynamics system of each pixel correspondence is by the FHN model description.
Step 2, neuron oscillator network oscillating layer, the coupling of each oscillator dynamic system model that modelling is got up according to FHN in the neural network is formed neural power network, and film potential V that i wherein is capable, j is listed as a neuron oscillator and voltage threshold R are respectively the rate of change of time;
Wherein, V is the potential difference (PD) of neuronal cell film both sides, is called film potential, and R is an internal state variable of representing the current potential threshold value, V and the identical implication of R representative, I in the following formula
I, jRepresent the light stimulus that comes from the outside that i is capable, a j row neuron receives, equal the gray-scale value in the gray level image on the numerical value.
Step 3, attention mobility realize be defined in vibration adjustment α and the β of layer according to following formula (5) and (6), thereby the corresponding neuron pool of the object of realizing current attention is provided frequency and is raise, and the neuron pool of other object correspondences produced the inhibition signal, make it provide frequency and slow down, distinguish current concern object and other objects with this.
Wherein, (p, q) pixel of the capable q row of p in the presentation video, τ has a neuron at least in the moment of providing, τ-the 1st, and in the previous moment of τ, M (τ) is the neuronic quantity of moment τ at the state of granting, Δ (τ) is the neuronic set of moment τ at the state of granting, I
I, jRepresent the light stimulus that comes from the outside that i is capable, a j row neuron receives, equal the gray-scale value in the gray level image on the numerical value.
Described FHN model is the abbreviation of FitzHugh-Nagumo model.
Aforesaid visual pattern input layer before gray-scale value being input to neural power network, has carried out normalization, and the gray-scale value after the normalization is in [0,1] scope.
(V just has stable limit cycle and produces when being unsettled R), the value of α and β must satisfy formula (2) to have only the equilibrium point of working as
8(V
2+αβ-1)<0(2)
Perhaps formula (3)
8(V
2+αβ-1)>0
-(10V
2-9.2)>0°(3)
Aforesaid neuron oscillator network oscillating layer, (i, j) capable, the j row of the i in the presentation video, 1≤i≤M, 1≤j≤N in described formula (4);
Wherein, M and N are the wide and high of image; Δ V
I, jWith Δ R
I, jRepresent the influence of peripheral nerve unit, they are defined by following formula:
Δx
i,j=γ
i-1,j-1;i,j(x
i-1,j-1-x
i,j)+γ
i-1,j;i,j(x
i-1,j-x
i,j)+
γ
i-1,j+1;i,j(x
i-1,j+1-x
i,j)+γ
i,j-1;i,j(x
i,j-1-x
i,j)+
γ
i,j+1;i,j(x
i,j+1-x
i,j)+γ
i+1,j-1;i,j(x
i+1,j-1-x
i,j)+
(7)
γ
i+1,j;i,j(x
i+1,j-x
i,j)+γ
i+1,j+1;i,j(x
i+1,j+1-x
i,j)
Wherein x represents V or R.
Aforesaid attention mobility realizes layer, θ in following formula (9) and (10)
1>θ
2>0
f
1(x)=a
1x+b
1(a
1>0,b
1<0)(11)
f
2(x)=a
2x+b
2(a
2<0,b
2>0)(12)
The implementation process parameter is a
1=2, b
1=-4, a
2=-4, b
2=2, θ
1=0.1, θ
2=0.01, θ
α=0.8, θ
β=4.
A kind of method with the neuron network simulation attention mobility of the present invention and uses the method and the application thereof of FHN model to compare in the document in the past, has following remarkable advantages and useful effect:
1, the document different from the past of the coupling scheme between the neuron.
2, the attention mobility mechanism that the FHN model is used for the anthropomorphic dummy.
Description of drawings
Fig. 1 attention mobility system architecture synoptic diagram;
Fig. 2 is image and the corresponding synoptic diagram of neural power network;
Fig. 3 is the input picture synoptic diagram;
Fig. 4 is the successively concern object figure of method among the present invention to Fig. 3.
Embodiment
Below in conjunction with accompanying drawing the specific embodiment of the present invention is illustrated.
This system is divided into three parts, the visual pattern input layer, and neuron oscillator network oscillating layer, attention mobility realizes layer.Relation between three layers is shown in Fig. 1 attention mobility system architecture synoptic diagram.
This system at first sets up each pixel in the width of cloth visual pattern and all the neuron oscillators in the neural power network related one to one, allows this group neuron oscillator produce vibration then.Because the object that brightness is big is stronger to human eye retina's light stimulus, so I
InputThe increase increase that causes providing frequency, suppress and can produce other zones in the distribution process.
Single neuron by the FHN model description is arranged with the form of matrix and carried out model reconstruct, form neural power network.Each neuron oscillator neuron in all peripheral with it 8 fields except that the edge has coupled relation in the newly-established model, the neuron oscillator at edge according to the difference of its adjacent vibration generators respectively with 3 or 5 neuron oscillators couplings.The film potential V of the neuron oscillator that i is capable, j is listed as and voltage threshold R are respectively the rate of change of time in the neural power network:
(i, the pixel that j) i is capable in the presentation video, j is listed as are also represented the neuron oscillator that i is capable in the neural power network, j is listed as, 1≤i≤M, 1≤j≤N (M and N are the wide and high of image) in the equation.Δ V
I, jWith Δ R
I, jThe neuronic influence of 8 neighborhoods around the representative, they are defined by following formula:
Δx
i,j=γ
i-1,j-1,i,j(x
i-1,j-1-x
i,j)+γ
i-1,j;i,j(x
i-1,j-x
i,j)+
γ
i-1,j+1;i,j(x
i-1,j+1-x
i,j)+γ
i,j-1;i,j(x
i,j-1-x
i,j)+
γ
i,j+1;i,j(x
i,j+1-x
i,j)+γ
i+1,j-1;i,j(x
i+1,j-1-x
i,j)+
(7)
γ
i+1,j;i,j(x
i+1,j-x
i,j)+γ
i+1,j+1;i,j(x
i+1,j+1-x
i,j)
For the more frequent granting of the neuron of realizing corresponding obvious object, the neuron of other objects is with the lower frequency granting or do not provide such fact, we at first allow neural power Netowrk tape preset parameter α and β move, neuron up to corresponding same area is synchronous, and this also shows to have finished cuts apart task.Afterwards, whenever any neuron granting, its can produce two kinds of signals to own and other neurons: neuron own and that provide is together produced excitatoty signal, and what do not provide with it is the signal of inhibition.
In system when operation,, we can controlled variable α and the change of β, makes gathering way of α reduce big many of speed than β.The high frequency period oscillation phase will be jumped in the most significant like this zone, and other zones can be quiet relatively, notes also can transferring to the most significant zone.After receiving attention, this zone can be suppressed to allow other zones to become significantly.
We can draw from top analysis to equation (1), and α and β can control neuronic activity.
Can cause the change of peripheral neuron α and β parameter after the neuron granting, thereby realize the transfer of noting.Wherein the variation of α and β is defined by following equation:
Wherein
f
1(x)=a
1x+b
1(a
1>0,b
1<0)(11)
f
2(x)=a
2x+b
2(a
2<0,b
2>0)(12)
(τ has a neuron at least in the moment of providing for p, the q) pixel of the capable q row of p in the presentation video, and M (τ) is that τ is in the neuronic quantity of the state of granting constantly, and Δ (τ) is the neuronic set of moment τ at the state of granting, I
I, jThe light stimulus that comes from the outside that the expression neuron receives equals the gray-scale value in the gray level image on the numerical value.By a is set
1>0, b
1<0 and a
2<0, b
2>0, and function h
1, h
2, (i is j) to (p, q) signal of granting excitability or inhibition for the neuron during each is provided.
At first be to set up the visual pattern input layer.System according to formula (1) and formula (4) establishment, set up neural power network, and each neuron on the network and the pixel in the image set up relation one to one, and the gray-scale value of image also becomes the input value of neuromotor system, the I in the corresponding formula (1)
InputAccompanying drawing 2 is a neural power network and gray level image corresponding diagram one by one, each small circle is a pixel in the representative image both, an also neuron in the neural power network of representative image correspondence, stain is represented the neuron of denumerable number, and horizontal line is represented the connection between the neuron.
Next is a neuron oscillator network oscillating layer.By each neuron oscillator in the whole neural power network is found the solution with 4 rank Runge-Kutta methods, realize that the oscillator in the same object reaches synchronized oscillation, the oscillator between different objects then desynchronizes.
Be that attention mobility realizes layer at last.Change α and β according to the definition in formula (5), the formula (6), the result selects successively according to the object in the sized images of conspicuousness, and the granting frequency of corresponding neural oscillator can become big when the object of selecting was noted.
Next with the concrete image graph 3 explanation implementing procedures of a width of cloth.Fig. 3 is a width of cloth gray-scale map, forms by 4 targets, and wherein 1. be the sun, 2. corresponding grey scale value 255 is tree, and 3. corresponding grey scale value 195 is the chain of mountains, and 4. corresponding grey scale value 130 is sky, corresponding grey scale value 65.At first Fig. 3 is carried out the normalization operation, just pixel value span on it is become [0,1] from [0,255].Visual pattern after the normalization is input in the good dynamical system (4) of foundation, sets up good dynamical system and be meant according to formula (4) (5), the system that (6) are set up.Next utilize the classical Runge-Kutta method in 4 rank to find the solution to (4) formula, parameter alpha in the solution procedure=1, β=1.25 remain unchanged at first voluntarily, neuron oscillator in the regional area in neural power network reaches synchro-resonance, and solving result V and R obtain periodic solution in solution procedure.Realize that in attention mobility the parameter of using in the layer is respectively a
1=2, b
1=-4, a
2=-4, b
2=2, θ
1=0.1, θ
2=0.01, θ
α=0.8, θ
β=4.Pass through iterative, among Fig. 3 the granting frequency of 4 object correspondences according to 1. 2. 3. 4. order increase respectively, the granting frequency that when certain object increases, suppresses other objects, the granting frequency that is other 3 objects slows down, and realizes that thus 4. 3. 2. 1. attention mobility constantly switching between 4 objects.Pay close attention to successively shown in the object figure as Fig. 4 at the object of selecting successively among Fig. 3.
It should be noted that at last: above embodiment only in order to the explanation the present invention and and unrestricted technical scheme described in the invention; Therefore, although this instructions has been described in detail the present invention with reference to each above-mentioned embodiment,, those of ordinary skill in the art should be appreciated that still and can make amendment or be equal to replacement the present invention; And all do not break away from the technical scheme and the improvement thereof of the spirit and scope of invention, and it all should be encompassed in the middle of the claim scope of the present invention.
Claims (5)
1. method with the neuron network simulation attention mobility is characterized in that may further comprise the steps:
Step 1, visual pattern input layer are input to the gray-scale value of gray level image in the neural power network, there is one-to-one relationship in neuron oscillator on pixel in the image and the neural power network, and the neurodynamics system of each pixel correspondence is by the FHN model description;
Step 2, neuron oscillator network oscillating layer, the coupling of each oscillator dynamic system model that modelling is got up according to FHN in the neural network is formed neural power network, and film potential V that i wherein is capable, j is listed as a neuron oscillator and voltage threshold R are respectively the rate of change of time:
Wherein, V is the potential difference (PD) of neuronal cell film both sides, is called film potential, and R is an internal state variable of representing the current potential threshold value, V and the identical implication of R representative, I in the following formula
I, jThe light stimulus that comes from the outside that the expression neuron receives equals the gray-scale value in the gray level image on the numerical value;
Step 3, attention mobility realize be defined in vibration adjustment α and the β of layer according to following formula (2) and (3), thereby the corresponding neuron pool of the object of realizing current attention is provided frequency and is raise, and the neuron pool of other object correspondences produced the inhibition signal, make it provide frequency and slow down, distinguish current concern object and other objects with this;
Wherein, (p, q) pixel of the capable q row of p in the presentation video, τ has a neuron at least in the moment of providing, τ-the 1st, and in the previous moment of τ, M (τ) is the neuronic quantity of moment τ at the state of granting, Δ (τ) is the neuronic set of moment τ at the state of granting, I
I, jRepresent the light stimulus that comes from the outside that i is capable, a j row neuron receives, equal the gray-scale value in the gray level image on the numerical value;
Described FHN model is the abbreviation of FitzHugh-Nagumo model.
2. the method with the neuron network simulation attention mobility according to claim 1, it is characterized in that: described visual pattern input layer, before gray-scale value being input to neural power network, carried out normalization, gray-scale value after the normalization is in [0,1] scope.
3. the method with the neuron network simulation attention mobility according to claim 1 is characterized in that: have only that (V just has stable limit cycle and produces when being unsettled R), the value of α and β must satisfy formula (4) when equilibrium point
8(V
2+αβ-1)<0(4)
Perhaps formula (5)
8(V
2+αβ-1)>0
-(10V
2-9.2)>0°(5)。
4. the method with the neuron network simulation attention mobility according to claim 1 is characterized in that: described neuron oscillator network oscillating layer, (i, j) capable, the j row of the i in the presentation video, 1≤i≤M, 1≤j≤N in described formula (1);
Wherein, M and N are the wide and high of image; Δ V
I, jWith Δ R
I, jRepresent the influence of peripheral nerve unit, they are defined by following formula:
Δx
i,j=γ
i-1,j-1;i,j(x
i-1,j-1-x
i,j)+γ
i-1,j;i,j(x
i-1,j-x
i,j)+
γ
i-1,j+1;i,j(x
i-1,j+1-x
i,j)+γ
i,j-1;i,j(x
i,j-1-x
i,j)+
γ
i,j+1;i,j(x
i,j+1-x
i,j)+γ
i+1,j-1;i,j(x
i+1,j-1-x
i,j)+
(6)
γ
i+1,j;i,j(x
i+1,j-x
i,j)+γ
i+1,j+1;i,j(x
i+1,j+1-x
i,j)
5. the method with the neuron network simulation attention mobility according to claim 1 is characterized in that: described attention mobility realizes layer, θ in following formula (8) and (9)
1>θ
2>0
f
1(x)=a
1x+b
1(a
1>0,b
1<0)(10)
f
2(x)=a
2x+b
2(a
2<0,b
2>0)(11)
The implementation process parameter is a
1=2, b
1=-4, a
2=-4, b
2=2, θ
1=0.1, θ
2=0.01, θ
α=0.8, θ
β=4.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201010199227.5A CN101894295B (en) | 2010-06-04 | 2010-06-04 | Method for simulating attention mobility by using neural network |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201010199227.5A CN101894295B (en) | 2010-06-04 | 2010-06-04 | Method for simulating attention mobility by using neural network |
Publications (2)
Publication Number | Publication Date |
---|---|
CN101894295A true CN101894295A (en) | 2010-11-24 |
CN101894295B CN101894295B (en) | 2014-07-23 |
Family
ID=43103482
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201010199227.5A Active CN101894295B (en) | 2010-06-04 | 2010-06-04 | Method for simulating attention mobility by using neural network |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN101894295B (en) |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107179705A (en) * | 2017-05-25 | 2017-09-19 | 江西理工大学 | A kind of annular coupled oscillator system and method for realizing that explosion type is synchronous |
CN107909151A (en) * | 2017-07-02 | 2018-04-13 | 小蚁科技(香港)有限公司 | For realizing the method and system of notice mechanism in artificial neural network |
Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US6504871B1 (en) * | 1997-07-31 | 2003-01-07 | Lsi Logic Corporation | IDCT processor for use in decoding MPEG compliant video bitstreams meeting 2-frame and letterboxing requirements |
CN101447077A (en) * | 2008-12-18 | 2009-06-03 | 浙江大学 | Edge detection method of color textile texture image oriented to textile industry |
CN101587590A (en) * | 2009-06-17 | 2009-11-25 | 复旦大学 | Selective visual attention computation model based on pulse cosine transform |
-
2010
- 2010-06-04 CN CN201010199227.5A patent/CN101894295B/en active Active
Patent Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US6504871B1 (en) * | 1997-07-31 | 2003-01-07 | Lsi Logic Corporation | IDCT processor for use in decoding MPEG compliant video bitstreams meeting 2-frame and letterboxing requirements |
CN101447077A (en) * | 2008-12-18 | 2009-06-03 | 浙江大学 | Edge detection method of color textile texture image oriented to textile industry |
CN101587590A (en) * | 2009-06-17 | 2009-11-25 | 复旦大学 | Selective visual attention computation model based on pulse cosine transform |
Cited By (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107179705A (en) * | 2017-05-25 | 2017-09-19 | 江西理工大学 | A kind of annular coupled oscillator system and method for realizing that explosion type is synchronous |
CN107179705B (en) * | 2017-05-25 | 2019-12-31 | 江西理工大学 | Ring coupling oscillator system and method for realizing explosive synchronization |
CN107909151A (en) * | 2017-07-02 | 2018-04-13 | 小蚁科技(香港)有限公司 | For realizing the method and system of notice mechanism in artificial neural network |
CN107909151B (en) * | 2017-07-02 | 2020-06-02 | 小蚁科技(香港)有限公司 | Method and system for implementing an attention mechanism in an artificial neural network |
Also Published As
Publication number | Publication date |
---|---|
CN101894295B (en) | 2014-07-23 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
Grossberg et al. | Cortical dynamics of visual motion perception: short-range and long-range apparent motion. | |
Deco et al. | A neurodynamical cortical model of visual attention and invariant object recognition | |
Giszter et al. | Motor primitives and synergies in the spinal cord and after injury—the current state of play | |
JPH09505161A (en) | CNN bionic eye or other topographic sensory organs or combinations thereof | |
CN107224273B (en) | Central-peripheral nerve closed-loop rehabilitation training method and system based on optical brain imaging nerve feedback | |
Strauss et al. | Choice reaching with a LEGO arm robot (CoRLEGO): the motor system guides visual attention to movement-relevant information | |
CN105701540A (en) | Self-generated neural network construction method | |
CN101894295B (en) | Method for simulating attention mobility by using neural network | |
Hwang et al. | Predictive coding-based deep dynamic neural network for visuomotor learning | |
Cohen Duwek et al. | Biologically plausible spiking neural networks for perceptual filling-in | |
Behncke et al. | Cross-paradigm pretraining of convolutional networks improves intracranial EEG decoding | |
Varier et al. | Establishing, versus maintaining, brain function: a neuro-computational model of cortical reorganization after injury to the immature brain | |
Weng | A 5-chunk developmental brain-mind network model for multiple events in complex backgrounds | |
Lewis | Self-organization of locomotory controllers in robots and animals | |
DE102020126380A1 (en) | Neural network device, method of operation thereof, and neural network system containing the same | |
Levin | Technological Approach to Mind Everywhere (TAME): an experimentally-grounded framework for understanding diverse bodies and minds | |
CN101690687B (en) | Method for generating phosphene for simulating prothetic vision device | |
Pashaie et al. | Self-organization in a parametrically coupled logistic map network: A model for information processing in the visual cortex | |
Kuriyama et al. | Gradient Climbing Neural Cellular Automata | |
Dino | An Experimental Pedagogy of Concept Development in the Introductory Architectural Design Studio | |
Brosch et al. | The brain’s sequential parallelism: perceptual decision-making and early sensory responses | |
Gu et al. | A cortical network model for clinical EEG data analysis | |
Kappen | Stimulus-dependent correlations in stochastic networks | |
CN107341543A (en) | A kind of cerebellar model modeling method based on intensified learning | |
Bálya et al. | A realistic mammalian retinal model implemented on complex cell CNN universal machine |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
C14 | Grant of patent or utility model | ||
GR01 | Patent grant | ||
TR01 | Transfer of patent right | ||
TR01 | Transfer of patent right |
Effective date of registration: 20230918 Address after: Room 603, 6th Floor, Building 6, Yard A1, Guanghua Road, Tongzhou District, Beijing, 101113 Patentee after: Beijing Hongzhou Culture Co.,Ltd. Address before: 100124 No. 100 Chaoyang District Ping Tian Park, Beijing Patentee before: Beijing University of Technology |