https://github.com/naveenkumar48/imageclassification

CNN-based image classifier for dogs and cats using TensorFlow and Keras, trained on the Kaggle dataset with automated preprocessing, training, and prediction.
https://github.com/naveenkumar48/imageclassification

cnn computer-vision deep-learning keras tensorflow

Last synced: 2 months ago
JSON representation

CNN-based image classifier for dogs and cats using TensorFlow and Keras, trained on the Kaggle dataset with automated preprocessing, training, and prediction.

• Host: GitHub
• URL: https://github.com/naveenkumar48/imageclassification
• Owner: Naveenkumar48
• Created: 2025-07-05T03:05:24.000Z (about 1 year ago)
• Default Branch: main
• Last Pushed: 2025-07-05T04:11:34.000Z (about 1 year ago)
• Last Synced: 2025-07-05T04:24:05.110Z (about 1 year ago)
• Topics: cnn, computer-vision, deep-learning, keras, tensorflow
• Language: Jupyter Notebook
• Homepage: https://www.kaggle.com/datasets/salader/dogs-vs-cats
• Size: 172 KB
• Stars: 0
• Watchers: 0
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

README

          
# 🐶🐱 Dog vs Cat Image Classifier using CNN

A deep learning project to automatically classify images of dogs and cats using a Convolutional Neural Network (CNN) built with TensorFlow and Keras.

---

## Project Overview

This project demonstrates an end-to-end workflow of:

- Downloading and preparing image data (from Kaggle).

- Building and training a CNN for binary classification.

- Evaluating performance with loss and accuracy metrics.

- Predicting and visualizing the classification of new images.

---

##Key Features

✅ Download dataset automatically via Kaggle API  

✅ Preprocess and normalize images  

✅ Build a custom CNN with multiple convolutional and dense layers  

✅ Visualize training and validation metrics  

✅ Predict on custom test images  

---

## Model Architecture

- **Input:** 256x256 RGB images

- **Layers:**

  - 3 × Conv2D layers with ReLU activation

  - MaxPooling and BatchNormalization after each Conv2D

  - Flatten + Dense layers with Dropout regularization

  - Sigmoid output for binary classification

- **Optimizer:** Adam

- **Loss Function:** Binary Crossentropy

- **Metrics:** Accuracy

---

## 🛠️ Technologies Used

- Python 3

- TensorFlow

- Keras

- OpenCV

- Matplotlib

- Kaggle API

- Jupyter Notebook / Google Colab

## 📈 Training & Evaluation

The model is trained over **5 epochs** on the Kaggle "Dogs vs Cats" dataset.  

### Sample Training Results:

- **Training Accuracy:** ~85% (depending on epochs and dataset splits)

- **Validation Accuracy:** ~80%

### Loss and Accuracy Graphs:

  

![Accuracy](Accuracy_output.png)

  

![Loss](Loss_output.png)

## 🖼️ Example Prediction

Test image loaded with OpenCV:

![Test Image](dog.jpg)

Model output: **Dog**