Image Classification Web Application

Table of Contents 📑

• Overview
• Steps to Use the Application
• Model Used
• Installation
• Code and Resources Used

Overview

A simple web demo with minimal framework using PyTorch and Streamlit to showcase an image classification model

Steps to Use the Application

• Clone the repository onto your own local machine

• Open command prompt/terminal

• Run pip install -r requirements.txt

• Type ‘streamlit run steamlit_ui.py’ in the command prompt/terminal

• A localhost address should automatically open in your web browser. If not, copy the local URL from the command prompt/terminal into your web browser.

• Click 'Browse files' and upload an image file in jpg format

• Now just wait for the results. 5 predictions should be outputted from highest probability to lowest probability.

Model Used

ResNet-101

ResNet-101 is a convolutional neural network that is 101 layers deep. You can load a pretrained version of the network trained on more than a million images from the ImageNet database. The pretrained network can classify images into 1000 object categories, such as keyboard, mouse, pencil, and many animals.

Installation

pip install -r requirements.txt

Code and Resources Used

• Python: Version 3.7.4
• Packages: PIL, torchvision, torch, streamlit