400 Birds Species Classification

🎦 Demo 👇

Self-written Medium blog: 400 Birds Species Classification

Aim of this Project

• This project aims to predict revenue for a set of customers for a time period in the future.
• Different machine learning & deep learning techniques were used.
• Deployment was done using Flask API on an AWS EC2 instance.

Datasets:-

• Source ➡️ Kaggle

📝 Best Model

• VGG19 with fne tuning of last few layers by freezing first 17 layers resulted in lowest loss of 0.14 on completely unseen test images.

Deployment using Flask API

• A simple web-app has been built using this model (as shown in the Demo).
• This web-app has also been Deployed using Flask API.
• Using this web-app, you can upload an image of a bird and the web-app will return the species name in the UI.

📁 Libraries Used

🖍️ scikit-learn 🖍️ Keras 🖍️ Tensorflow 🖍️ matplotlib 🖍️ seaborn 🖍️ numpy 🖍️ pandas 🖍️ prettytable 🖍️ Flask

🛠️ 🧰 Tools and Softwares Used

• Google Colab
• Jupyter Notebook
• Sublime Text