dog-breed-classifier

A machine learning model that identifies dog breeds from images with high accuracy (~99% validation accuracy).

Requirements

• Python 3.8+ (tested with Python 3.12)
• Dataset of 10 dog breeds (see Dataset section)
• 4GB+ RAM (no GPU required)

Dataset

Download the dataset from: Kaggle - Dog Breed Image Dataset

Place the extracted dataset in the model/ directory with the following structure:

dataset/
  ├── Beagle/
  ├── Boxer/
  ├── Bulldog/
  ├── Dachshund/
  ├── German_Shepherd/
  ├── Golden_Retriever/
  ├── Labrador_Retriever/
  ├── Poodle/
  ├── Rottweiler/
  └── Yorkshire_Terrier/

Install requirements

pip install -r requirements.txt

Train and build model

python3.12 model.py

Output: Creates a trained model file dog_breed_classifier.keras and performance visualizations.

Predict dog breed in CLI

python3.12 model.py <image_path>

Example:

python3.12 model.py test_dog.jpg

Predict dog breed and create a plot

python3.12 predict_with_plots.py

Output: Creates a visualization showing the predicted breed and confidence scores.

Model Performance

• Validation Accuracy: ~99%
• Supported Breeds: 10 popular dog breeds
• Processing Time: ~1 second per image on CPU