Convolutional Neural Networks for Images
CNNs are the workhorse for vision and introduce the idea of learned spatial feature hierarchies.
Learn it
Read the CS231n convolutional networks notes for architecture intuition (filters, stride, pooling), then build an image classifier end-to-end with the official Keras 'Image classification from scratch' example on the Cats vs. Dogs dataset, including augmentation and dropout.
By the end you’ll be able to
- Explain convolution, pooling, and feature-map hierarchies
- Build and train a CNN image classifier
Free resources
Practice
Train the Keras from-scratch CNN, then tweak filter counts and augmentation and record the effect on validation accuracy.
Build it — your output
A trained CNN image classifier with a documented accuracy on a held-out set.
Prove it — check yourself
You’re ready to move on when you can honestly say yes to each:
- Why does a convolutional layer use far fewer parameters than a dense one?
- What does a pooling layer accomplish?
- How do early vs. late layers differ in what they detect?
Do it with AI
Use a free AI tool as your tutor and pair-programmer. Copy a prompt below into Claude, ChatGPT, or Gemini (for building, try Cursor or Copilot).
The habit: Try it yourself first — a little struggle is where the learning actually happens. Then ask for a hint before the answer — never copy blindly.
- Understand itClaude / ChatGPT / Gemini
I'm a beginner learning "Convolutional Neural Networks for Images". Explain it in simple terms with a real-world analogy, then ask me ONE question to check my understanding — don't reveal the answer.
- PracticeClaude / ChatGPT / Gemini
Give me 3 short practice problems on "Convolutional Neural Networks for Images" (my goal: explain convolution, pooling, and feature-map hierarchies), from easy to hard. Don't include the solutions yet — I'll try them first, then ask you to check.
- Get unstuckClaude / ChatGPT / Gemini
I'm stuck on this exercise about "Convolutional Neural Networks for Images": [paste your attempt here]. Give me ONE small hint to move forward — do NOT give me the full solution.
- Build & reviewCursor / Copilot
Here is my work for a "Convolutional Neural Networks for Images" task: [paste your code or answer]. Review it: point out what's wrong and WHY, and suggest improvements — but let me make the fixes myself.
- Prove itClaude / ChatGPT / Gemini
Quiz me on "Convolutional Neural Networks for Images". Ask me 5 questions one at a time, wait for my answer each time, then tell me if I'm right and briefly why.