Deep Learning · TEE · guide

CNN architectures and transfer learning

  1. AlexNet and the reason for its components
  2. Inception examines several spatial scales
  3. Why one by one convolution saves work
  4. ResNet adds a shortcut
  5. Transfer learning from a small dataset
  6. Model answer: feature extraction and fine-tuning
  7. How to draw the architecture diagrams
  8. Inception concatenation and residual addition
  9. Model answer: AlexNet, Inception and ResNet
  10. Applications of convolutional neural networks
  11. Reading LeNet and AlexNet layer diagrams
  12. Top-1 and top-5 error: what an ImageNet result means
  13. VGG16: small filters in a regular sequence
  14. Global average pooling: one average for each channel
  15. The faculty's 480-channel bottleneck numerical
  16. Batch normalization inside a residual network
  17. Residual-block order and architecture qualifications
  18. Choosing frozen layers from dataset size and similarity
  19. Four ways to reuse a pretrained model