News classification
Contact us
- Add: No. 9, North Fourth Ring Road, Haidian District, Beijing. It mainly includes face recognition, living detection, ID card recognition, bank card recognition, business card recognition, license plate recognition, OCR recognition, and intelligent recognition technology.
- Tel: 13146317170 廖经理
- Fax:
- Email: 398017534@qq.com
Image recognition and neural networks
Image recognition and neural networks
Image recognition and neural networks
1. Definition of convolutional neural networks
Convolutional Neural Networks (CNN) is a kind of feedforward neural network with convolution or correlation calculation and deep structure. It is one of the representative algorithms of deep learning. The CNN consists of different convolutional and pooling layers. It is widely used in computer vision. For example, image classification, object recognition, action recognition, pose estimation, and neural style transfer.
2. The working process of convolutional neural networks
The convolutional neural network will perform multiple samples and record multiple features of the object multiple times. In addition to these connection layers, there are pooling and convolution layers. CNN preserves important feature information in image recognition while also reducing the size of the input. The input and output of the convolutional layer are all multiple matrices. The convolutional layer contains a plurality of convolution kernels, each convolution kernel is a matrix, and each convolution kernel is equivalent to a filter, which can output a specific feature map, and each feature map is a convolution layer. An output unit.
for example. The computer is going to do image processing. First read the picture. Computers don't capture all the features of an object in seconds, just like a human. It needs to be understood. For ease of understanding, it converts each image into a series of specifically ordered points (pixels). If you change the order or color of the pixels, the image changes. The computer will attempt to extract features from the image by using the arrangement of the image's space. In order to understand the image, understanding how pixels are arranged is extremely important for a network. This is what the convolutional network has to do.
(2) Reduce the number of parameters of the image. The CNN combines features of similar meaning into the same feature and merges the adjacent features to a closer position. Since the relative position of each feature forming a specific theme may be slightly changed, the position of the strongest intensity in the feature map may be input by sampling, and the dimension of the intermediate representation (ie, the size of the feature map) is reduced, thereby even localizing The feature has a certain degree of displacement or distortion, and the model can still detect this feature.
1. Definition of convolutional neural networks
Convolutional Neural Networks (CNN) is a kind of feedforward neural network with convolution or correlation calculation and deep structure. It is one of the representative algorithms of deep learning. The CNN consists of different convolutional and pooling layers. It is widely used in computer vision. For example, image classification, object recognition, action recognition, pose estimation, and neural style transfer.
2. The working process of convolutional neural networks
The convolutional neural network will perform multiple samples and record multiple features of the object multiple times. In addition to these connection layers, there are pooling and convolution layers. CNN preserves important feature information in image recognition while also reducing the size of the input. The input and output of the convolutional layer are all multiple matrices. The convolutional layer contains a plurality of convolution kernels, each convolution kernel is a matrix, and each convolution kernel is equivalent to a filter, which can output a specific feature map, and each feature map is a convolution layer. An output unit.
for example. The computer is going to do image processing. First read the picture. Computers don't capture all the features of an object in seconds, just like a human. It needs to be understood. For ease of understanding, it converts each image into a series of specifically ordered points (pixels). If you change the order or color of the pixels, the image changes. The computer will attempt to extract features from the image by using the arrangement of the image's space. In order to understand the image, understanding how pixels are arranged is extremely important for a network. This is what the convolutional network has to do.
3. Benefits of using convolutional networks
(2) Reduce the number of parameters of the image. The CNN combines features of similar meaning into the same feature and merges the adjacent features to a closer position. Since the relative position of each feature forming a specific theme may be slightly changed, the position of the strongest intensity in the feature map may be input by sampling, and the dimension of the intermediate representation (ie, the size of the feature map) is reduced, thereby even localizing The feature has a certain degree of displacement or distortion, and the model can still detect this feature.