Animal Recognition using VGG16

Introduction

This project is a part of the Habib University's CS 351 Artificial Intelligence coursework. The goal of this project is to build a convolutional neural network (CNN) that can classify images of animals. The dataset used for this project is the Animals-10 dataset from Kaggle. The dataset contains 10 classes of animals with 1300 images per class. The classes are as follows:

• dog
• cat
• wild bird
• chicken
• sheep
• cow
• elephant
• bear
• spider
• squirrel

Methodology

The methodology used for this project is as follows:

1. Data Preprocessing
2. Model Building
3. Model Training
4. Model Evaluation

Data Preprocessing

The data preprocessing step involves the following steps:

1. Loading the dataset
2. Splitting the dataset into training and testing sets
3. Resizing the images to 224x224
4. Normalizing the images
5. One-hot encoding the labels

Model Building

The model building step involves the following steps:

1. Loading the VGG16 model
2. Removing the last layer of the VGG16 model
3. Adding a new layer to the VGG16 model
4. Freezing the weights of the VGG16 model
5. Compiling the model

Model Training

The model training step involves the following steps:

1. Training the model
2. Saving the model

Model Evaluation

The model evaluation step involves the following steps:

1. Loading the model
2. Evaluating the model

Results

The model achieved an accuracy of 92.5% on the testing set.

Acknowledgements

• Animals-10 dataset
• VGG16 model
• Dr. Saleha Raza

Authors

• Ali Asghar Chakera
• Abbas Haider
• Umema Zehra