Practice Flashcards
Flip to reveal answersWhat does convolution do in a CNN?
Track your progress — Sign up free to save your progress and get smart review reminders based on spaced repetition.
All 5 Flashcards — CNNs
Sign up free to track progress and get spaced-repetition review schedules.
Question
What does convolution do in a CNN?
Answer
Slides a small **filter** of weights across the input, multiplying and summing at each position to produce a **feature map** of where that pattern occurs.
Question
Why are shared weights important?
Answer
The **same** filter is used at every position, so a feature is recognised **wherever it appears** — and one detector costs a handful of weights rather than a set per location.
Question
What is pooling for?
Answer
Keeping the strongest value in each block: **fewer numbers** to carry forward, and tolerance to small shifts in position.
Question
What do successive CNN layers learn?
Answer
**Edges**, then shapes built from edges, then objects built from shapes — a hierarchy that is learned rather than designed.
Question
Why is flattening an image a problem?
Answer
It **destroys the geometry** — the network no longer knows which pixels were adjacent — and a 200×200 colour image gives 120,000 inputs, needing millions of weights.
Read the notes
Full study notes for CNNs
Topic 4.3 hub
Machine learning approaches
More from Topic 4.3
All flashcards in this topic
Computer Science exam skills
Paper structures & tips
Track your progress with spaced repetition
Sign up free — Aimnova tells you exactly which cards to review and when, so you remember everything before your IB exam.
Start Free