ausjahk/flower-classifier-from-scratch

Flower Species Classifier (CNN From Scratch)

Overview

Custom Convolutional Neural Network (CNN) built and trained entirely from scratch (random weight initialization) on the Kaggle Flowers Recognition dataset.

Classes (5)

• Daisy 🌼
• Dandelion 🌾
• Rose 🌹
• Sunflower 🌻
• Tulip 🌷

Architecture

• 4 Conv Blocks (32 → 64 → 128 → 256 filters)
• BatchNorm + ReLU + MaxPool2d
• Dense Classifier Head (256 -> 5 classes)
• Total Parameters: 4,585,221 (All random initialization)

Performance

• Test Accuracy: 64.35%