Skip to content
AI-grafen
Sign in
☰ Menu

The goal E University

Image classification with convolutional networks

Build a CNN, understand convolution and receptive fields, and look inside the network to see what it has actually learned.

Knowledge nodes
48
From zero
about 37 h
Labs
6
See what you already know — no account

The diagnostic removes what you already know, so your path is usually much shorter.

What you can do afterwards

Labs along the way

You write the code. Tests you cannot see decide whether it holds up.

Lab: a neural network in pure NumPy — forward, backprop, gradient checkDa sandbox · about 75 minLab: dot product, norm and cosine similarityDin the browser · about 40 minLab: gradient descent from scratchDin the browser · about 50 minLab: matrix multiplication and one layer of a neural networkDin the browser · about 45 minLab: tensors, autograd and your first training loopDa sandbox · about 50 minLab: train a digit classifier in PyTorchDa sandbox · about 60 min

The whole path

Everything the goal builds on, grouped by level and in the order it builds on itself. Show on the map

AExplorer3 knowledge nodes
  1. Patterns and categories
  2. Sequences and precise instructions
  3. Integer arithmetic and order of operations
BInvestigator6 knowledge nodes
  1. Data in everyday life
  2. Algorithmic thinking
  3. Rule-based systems and machine learning
  4. Binary numbers and bits
  5. Training data, features and labels
  6. Classification: how a model sorts information
CBuilder9 knowledge nodes
  1. Functions and coordinate systems
  2. Programming logic — variables, conditions, loops
  3. Python — the basics
  4. Python — lists, loops and dictionaries
  5. Statistics — mean, median and spread
  6. Digital representation: text, images and audio
  7. Linear regression: fitting a straight line to data
  8. Neural networks — the intuition
  9. Images as matrices
DAI developer17 knowledge nodes
  1. Derivatives and optimisation
  2. Gradient descent
  3. Vectors
  4. Matrices and matrix multiplication
  5. Convolution by hand
  6. Linear regression with several features
  7. Neural networks — the forward pass with matrices
  8. Activation functions
  9. Backpropagation
  10. What does the network see? Images through the layers
  11. NumPy — arrays and vectorisation
  12. Overfitting and generalisation
  13. PyTorch — tensors and autograd
  14. Dataloaders, batches and epochs
  15. Training, validation and test
  16. Train a neural network in PyTorch
  17. MNIST from scratch
EUniversity11 knowledge nodes
  1. Convolutional networks (CNNs)
  2. Vanishing and exploding gradients
  3. Image preprocessing and augmentation
  4. Residual connections
  5. CNN architectures: LeNet to ResNet
  6. Computer vision — the basics
  7. Regularisation: dropout, weight decay, early stopping
  8. Data augmentation in practice
  9. Transfer learning
  10. Image classification end to end
  11. Visualising what a CNN sees
FAI engineering2 knowledge nodes
  1. Object detection
  2. Segmentation