FINAL YEAR PROJECT 2023 (RE-CREATION)

About

• The main objective of this project is to classify between normal and pneumonia affected chest x-rays.
• This project is the reconstruction of my final year project for the Bachelor of Science in Computer Science.
• During the initial development for the final year project work assessment, Transfer Learning technique was not considered for the Model Building process.
• Through experimentation, I discovered that Transfer Learning is a great choice. I decided to use the pre-trained weights for DenseNet-121 trained on a vast ImageNet Dataset.
• I used pre-trained DenseNet-121 weights from the ImageNet dataset as a feature extraction layer of the new model for the faster development to the solution. I used the newly built model to train it on the labeled chest x-ray dataset for the detection of pneumonia in chest x-rays.
• The primary goal is to incorporate the recommendations provided by the esteemed Examiner after the presentation of the original final year project.

Try it

• Clone this repository

git clone https://github.com/dipakexe/FINAL_YEAR_PROJECT_2023.git

• Build the frontend

npm run build

• Run the server

python app.py

It will automatically download the the pre-trained model from here and save it tomodels/ directory. Then It will start the server.