This project is a Dog Breed Classification model developed using TensorFlow 2.0 and Python. The model utilizes Convolutional Neural Networks (CNNs) and Transfer Learning to accurately identify the breed of dogs from images. This application of deep learning showcases the power of CNNs in image classification tasks and the efficiency of Transfer Learning for leveraging pre-trained models to achieve high accuracy with relatively little data.
- TensorFlow 2.0: Utilizes the latest features of TensorFlow for model building and training.
- Transfer Learning: Employs a pre-trained model as the foundation to improve learning efficiency and accuracy.
- CNNs: Implements Convolutional Neural Networks to automatically and effectively extract features from dog images.
- Data Preprocessing: Uses pandas for data manipulation and preprocessing to prepare the dataset for training.
- Visualization: Leverages matplotlib for visualizing training results, including accuracy and loss over epochs.
The dataset used for training the Dog Breed Classification model consists of thousands of labeled images of dogs from various breeds. It is structured to train the model effectively by using a balanced number of images for each breed to avoid bias.
The core of the model is a pre-trained CNN (specify which one, e.g., ResNet50, InceptionV3, etc.) on top of which additional layers are added for the task of dog breed classification. The model's architecture is designed to efficiently learn from the dog images dataset and accurately classify dog breeds.
To run this project on your local machine:
git clone https://github.com/Kryp6405/dog-breed-classification.git