Deep Learning - Convolutional Neural Networks – Tiven's Blog

Tiven Wang

CNN TensorFlow Deep Learning Python

Deep Learning - Convolutional Neural Networks

Wang Tiven July 11, 2018

425 favorites

bookmark

share

Filters and Feature Maps
Building blocks
词汇

想要深入理解卷积神经网络，就要理解数学里卷积 (Convolution) 的概念，可以参见 Convolution，还可以阅读 Chris Olah’s blog post - Understanding Convolutions

Filters and Feature Maps

\[m_{ij}^k = f((W * x)_{ij} + b^k)\]

deep convolutional networks

Building blocks

Convolutional layer

神经元可以表示为

\[y = \sigma(Wx + b)\]

Pooling layer

ReLU layer

Fully connected layer

Loss layer

词汇

locally invariant 局部不变
fractional max pooling 分数最大池

References

[1] CS231n Convolutional Neural Networks for Visual Recognition http://cs231n.github.io/convolutional-networks/
[2] Towards Data Science - Cutting-Edge Face Recognition is Complicated. These Spreadsheets Make it Easier. https://towardsdatascience.com/cutting-edge-face-recognition-is-complicated-these-spreadsheets-make-it-easier-e7864dbf0e1a

Similar Posts

TensorFlow 2 - Serving with Docker One of the easiest ways to get started using TensorFlow Serving is with Docker.
TensorFlow 2 - quickstart for beginners Start our journey with TensorFlow
Deep Learning - Regularization Regularization for Deep Learning
Python - pandas pandas
Machine Learning - Keras Keras
TensorFlow TensorFlow

Comments

Back to Top