Cortex AI logo placeholderCortexLabAI · learning · engineering
Academy/Deep Learning Systems/CNN
module 2 of 7 · 40 min

CNN

Learn local spatial patterns with shared convolutional filters.

Explain CNN clearlyImplement a small CNN exampleEvaluate whether CNN improves a simpler baselineIdentify failure cases and operational constraints
learning statenot started
0% completesign in to track progress
mental model

start with the idea before the implementation.

A convolution scans the same learned detector across spatial positions.
core concepts

the mechanisms you need to reason about.

01

Kernels

Kernels is studied through intuition implementation evidence and trade-offs. The goal is to be able to explain the mechanism and verify it with a concrete test rather than only repeat a definition.

02

Receptive fields

Receptive fields is studied through intuition implementation evidence and trade-offs. The goal is to be able to explain the mechanism and verify it with a concrete test rather than only repeat a definition.

03

Pooling

Pooling is studied through intuition implementation evidence and trade-offs. The goal is to be able to explain the mechanism and verify it with a concrete test rather than only repeat a definition.

04

Augmentation

Augmentation is studied through intuition implementation evidence and trade-offs. The goal is to be able to explain the mechanism and verify it with a concrete test rather than only repeat a definition.

engineering lab

turn the lesson into evidence.

LAB 1

train a small image classifier

Build the smallest version first. Record the input, expected output, measured result and one failure you discovered.

LAB 2

inspect activation maps

Build the smallest version first. Record the input, expected output, measured result and one failure you discovered.

LAB 3

test augmentation

Build the smallest version first. Record the input, expected output, measured result and one failure you discovered.

knowledge checks

prove you can explain and decide.

1

explain parameter sharing

ask cortex to test me →
2

calculate output shape

ask cortex to test me →
3

identify overfitting

ask cortex to test me →
failure modes

what usually goes wrong.

risk

data leakage through augmentation

Detect this early by defining a baseline, a measurable signal and a condition that would cause you to stop or redesign the approach.

risk

oversized models

Detect this early by defining a baseline, a measurable signal and a condition that would cause you to stop or redesign the approach.

risk

ignoring class imbalance

Detect this early by defining a baseline, a measurable signal and a condition that would cause you to stop or redesign the approach.

proof of learning

Create a short CNN engineering note with one working artifact one metric one failure case and one decision about when you would or would not use it.

Save the result in your portfolio or project repository. A strong learning artifact should make your assumptions, metrics and failure analysis visible.