This is a small project to recognize fruits using 3 popular CNN models: ResNet50, DenseNet121 and MobileNet. These models are trained on our custom dataset of 20 fruits. The dataset is available here. The website is built using Streamlit and deployed on Streamlit Cloud. You can access the website here.

If you want to run this project locally, you can clone this repository and install the required packages using the following command:

# Clone the repository
git clone https://github.com/vm7608/fruit-classification-website.git

# Create a virtual environment
virtualenv .venv

# Activate the virtual environment
source .venv/bin/activate

# Install the required packages
pip install -r requirements.txt

# Run the app
streamlit run streamlit_app.py

We have created a full training/testing pipeline for the models. The pipeline details can be found in the folder notebooks. The trained models are saved in the folder models. The code to process data and create the dataset can be found in the folder data processing. The code to create the website can be found in the folder streamlit_app.

For more details, please refer to the document folder.

• Cao Kieu Van Manh
• Nguyen Tuan Hung
• Vo Hoang Bao
• Nguyen Tien Hung