Pattern recognition method and device
Pattern recognition method and device
 CN 104,794,501 B
 Filed: 05/14/2015
 Issued: 01/05/2021
 Est. Priority Date: 05/14/2015
 Status: Active Grant
First Claim
1. A pattern recognition method, comprising:
 receiving data to be identified, wherein the data to be identified is a picture;
performing pattern recognition on the data to be recognized by using a pattern recognition model obtained by training a convolutional neural network with a recursive convolutional layer;
wherein, the convolutional neural network with the recursive convolutional layer is a neural network which combines recursive input in the convolutional layer on the basis of feedforward input to obtain total input and carries out nonlinear excitation on the total input,the feed forward input is specifically represented as;
wherein u is^{(i,j)}Representing the response of a local cell centered at (i, j) on the vectorized priorlayer feature map,representing weights of the vectorized feedforward input;
the recursive input in the convolution layer is specifically expressed as;
wherein x is^{(i,j)}(t1) represents the response of the local cell centered at (i, j) on the current layer feature map at the previous time step,representing the weight of vectorized intralayer recursive input, T representing transposition, T representing a time value, and T being less than or equal to N, wherein N represents the total time step number;
the pattern recognition comprises any one of;
face recognition, gesture recognition and traffic sign recognition.
Chinese PRB Reexamination
Abstract
The invention provides a pattern recognition method and a device, wherein the pattern recognition method comprises the following steps: receiving data to be identified; performing pattern recognition on the data to be recognized by using a pattern recognition model obtained based on convolutional neural network training with a recursive convolutional layer; the convolutional neural network with the recursive convolutional layer is a neural network which combines recursive inputs in the layer on the basis of feedforward input to obtain total input and performs nonlinear excitation on the total input. The convolutional neural network with the recursive convolutional layer can be fully integrated with context information, and the depth of the network is increased under the condition that the quantity of parameters is kept unchanged, so that the accuracy of pattern recognition is effectively improved.
6 Claims

1. A pattern recognition method, comprising:

receiving data to be identified, wherein the data to be identified is a picture; performing pattern recognition on the data to be recognized by using a pattern recognition model obtained by training a convolutional neural network with a recursive convolutional layer; wherein, the convolutional neural network with the recursive convolutional layer is a neural network which combines recursive input in the convolutional layer on the basis of feedforward input to obtain total input and carries out nonlinear excitation on the total input, the feed forward input is specifically represented as; wherein u is^{(i,j)}Representing the response of a local cell centered at (i, j) on the vectorized priorlayer feature map,representing weights of the vectorized feedforward input; the recursive input in the convolution layer is specifically expressed as; wherein x is^{(i,j)}(t1) represents the response of the local cell centered at (i, j) on the current layer feature map at the previous time step,representing the weight of vectorized intralayer recursive input, T representing transposition, T representing a time value, and T being less than or equal to N, wherein N represents the total time step number; the pattern recognition comprises any one of; face recognition, gesture recognition and traffic sign recognition.


2. The method of claim 1, wherein the total input is formulated as:
wherein z is_{ijk}(t) denotes the total input, b_{k}Is a deviation value.

3. The method of claim 2, wherein nonlinear excitation of the total input is formulated as:
x_{ijk}(t)＝
f(z_{ijk}) (t) wherein x_{ijk}(t) represents the response of the current time step unit, and f is the nonlinear excitation function.

4. A pattern recognition apparatus, comprising:

the device comprises a receiving unit, a judging unit and a judging unit, wherein the receiving unit is used for receiving data to be identified, and the data to be identified is a picture; the pattern recognition unit is used for carrying out pattern recognition on the data to be recognized by utilizing a pattern recognition model obtained by training a convolutional neural network with a recursive convolutional layer; wherein, the convolutional neural network with the recursive convolutional layer is a neural network which combines recursive input in the convolutional layer on the basis of feedforward input to obtain total input and carries out nonlinear excitation on the total input, the feed forward input is specifically represented as; wherein u is^{(i,j)}Representing the response of a local cell centered at (i, j) on the vectorized priorlayer feature map,representing weights of the vectorized feedforward input; the recursive input in the convolution layer is specifically expressed as; wherein x is^{(i,j)}(t1) represents the last oneTimestep response of local cell centered at (i, j) on current layer feature map,representing the weight of vectorized intralayer recursive input, T representing transposition, T representing a time value, and T being less than or equal to N, wherein N represents the total time step number; the pattern recognition comprises any one of; face recognition, gesture recognition and traffic sign recognition.


5. The apparatus of claim 4, wherein the total input is formulated as:
wherein z is_{ijk}(t) denotes the total input, b_{k}Is a deviation value.

6. The apparatus of claim 5, wherein the nonlinear excitation of the total input is formulated as:
x_{ijk}(t)＝
f(z_{ijk}) (t) wherein x_{ijk}(t) represents the response of the current time step unit, and f is the nonlinear excitation function.
Specification(s)