Deep Learning with Pytorch in a Nutshell
  • Content
  • Image recognition
  • Object detection
  • Semantic segmentation
  • GAN
  • Image style transfer
  • Face recognition
  • Interpretability
  • Word embedding
  • Pytorch
  • Optimization
  • Special layers
    • Transposed convolution
  • Neural architecture search
  • Reinforcement learning
    • Proof of Bellman equation
    • Tabular solution method
Powered by GitBook
On this page
  • Network in network (NIN)
  • Very Deep Convolutional Networks for Large-Scale Image Recognition (VGG)
  • Deep Residual Learning for Image Recognition (Resnet)
  • Inception
  • Xception
  • Dense net
  • Squeeze and excitation network
  • Dual path network
  • Shuffle network
  • Squeeze network
  • ResXt
  • Fully Convolutional Attention Networks for Fine-Grained Recognition
  • Residual Attention Network for Image Classification

Image recognition

PreviousContentNextObject detection

Last updated 6 years ago

Network in network (NIN)

  • Global average pooling

  • 1*1 convolutional kernel

Very Deep Convolutional Networks for Large-Scale Image Recognition (VGG)

  • Simply stacks more convolution layers

  • Stack small convolutional kernels to achieve the same "receptive field" as bigger one.

    • Stack two 33 convolutional kernel is equivalent to 55 convolutional kernel

One 55 kernel and stacks two 33 kernels have the same effective receptive field.

Deep Residual Learning for Image Recognition (Resnet)

  • Add residual connection

Inception

Xception

Dense net

Squeeze and excitation network

Dual path network

Shuffle network

Squeeze network

ResXt

Fully Convolutional Attention Networks for Fine-Grained Recognition

Residual Attention Network for Image Classification

https://arxiv.org/pdf/1603.06765.pdf
https://arxiv.org/pdf/1704.06904.pdf
https://github.com/tengshaofeng/ResidualAttentionNetwork-pytorch
https://github.com/youansheng/AttentionModule
https://arxiv.org/abs/1312.4400
https://arxiv.org/abs/1409.1556
https://arxiv.org/abs/1512.03385