A convolutional neural network based
convolutional neural network functions: the problem of classification
convolution layer (image feature extraction) -> BN (normalized bulk) -> active layer (nonlinear features was introduced) -> Pooling layer (sample layer, to reduce the complexity of the image, wherein more concentrated) -> fully connected layers (integration features, classification);
Convolutional neural network infrastructure accumulation
Guess you like
Origin blog.csdn.net/alangaixiaoxiao/article/details/103136566
Recommended
Ranking