🔲 Pooling Layers
Reduce spatial dimensions while preserving important features
Max Pool
2×2 window, stride 2
→
Step 0 of 4
💡 Key Concepts
- Max Pooling: Takes the maximum value in each window (preserves strongest features)
- Average Pooling: Takes the average value (smooths features)
- 2×2, stride 2: Halves the spatial dimensions (4×4 → 2×2)
- Benefits: Reduces computation, provides translation invariance, reduces overfitting