Skip to content

Latest commit

 

History

History
88 lines (61 loc) · 2.94 KB

README.md

File metadata and controls

88 lines (61 loc) · 2.94 KB

🏆 Custom Image Classification Model for Animal Recognition

:dog, cow, cat, lamb, and zebra, each with 100 images sourced from the internet or captured using phone. Develop a classification model to classify these classes

Project Group No.: 1
Group Members:

  • Tausib Abrar (2121446642)
  • Fahim Muntashir (2021183642)
  • Md Abul Bashar Nirob (2022198042)
  • Muhammad Omar Mahin Jinnat (2012826642)

📌 Project Overview

This project aims to develop a custom image classification dataset featuring five animal classes:
🐶 Dog | 🐄 Cow | 🐱 Cat | 🐑 Lamb | 🦓 Zebra

We collected 100 images per class from online sources and personal captures, processed them, and developed a CNN model to classify these animals with an accuracy target of 90%+.

📂 Dataset Details

  • Total Images: 500 (100 per class).
  • Sources: Google Images, Unsplash, personal phone captures.
  • Preprocessing Applied:
    • Resized → Standardized to 224×224 pixels.
    • Normalized → Scaled pixel values to [0,1].
    • Augmented → Applied rotation, flipping, cropping to improve model generalization.

⚙️ Model Architecture & Tools

The model is a Convolutional Neural Network (CNN) designed for multi-class classification.

🔧 Libraries Used:

  • TensorFlow/Keras → Model Training
  • OpenCV → Image Preprocessing
  • NumPy & Pandas → Data Handling

🔍 Model Development Process:

  1. Simple CNN Model: Trained with 500 images (initial accuracy ~85%).
  2. Hyperparameter Tuning: Optimizing learning rate, batch size, epochs.
  3. Transfer Learning (Next Step): Testing with pre-trained models (VGG16, ResNet) for better performance.

🚀 How to Run the Code

Follow these steps to run the project:

🔹 Clone the Repository

git clone : (https://github.com/FahimMuntashir/animal-classifier.git)
cd custom-image-classification

🔹 Install Dependencies

pip install -r requirements.txt

🔹 Train the Model

python train.py

🔹 Evaluate the Model

python evaluate.py

📊 Results & Evaluation
	•	Initial CNN Model Accuracy: 85%
	•	Planned Enhancements:
	•	Use Transfer Learning (VGG16, ResNet)
	•	Further tuning of hyperparameters
	•	Increase dataset size for better generalization

🎥 Demo Video

📌 Watch the One-Minute Demo Here 

This demo showcases the model classifying images and its training progress.


📢 Future Work
	•	Experiment with deeper models (EfficientNet, MobileNet).
	•	Optimize preprocessing techniques to improve classification accuracy.
	•	Deploy the model as a web or mobile app for real-time classification.

📖 References
	1.	Krizhevsky, A., Sutskever, I., & Hinton, G. E. (2012). ImageNet Classification with Deep Convolutional Neural Networks.
	2.	TensorFlow Documentation: Image Classification Guide