Capsule Vision 2024 Challenge

Overview

The Capsule Vision 2024 Challenge involves developing AI models for the automatic classification of abnormalities in video capsule endoscopy (VCE) video frames. The goal is to create a vendor-independent and generalized AI model capable of classifying ten different types of abnormalities.

This repository contains the code, documentation, and resources for our participation in the Capsule Vision 2024 Challenge. The project includes data preprocessing, model training, evaluation, and submission scripts.

Team

• Utsav Singhal: Main Lead and Project Coordinator
• Serra Aksoy: Training and Evaluating Models
• Anshul Bansal: Data Preprocessing

Project Structure

The repository is organized as follows:

capsule-vision-2024-challenge/
├── data/               # Data files (Training, Validation, Testing datasets)
├── notebooks/          # Jupyter notebooks for exploration and experimentation
├── src/                # Source code for data preprocessing, model training, and evaluation
│   ├── data_preprocessing.py  # Scripts for data cleaning and preprocessing
│   ├── model_training.py      # Scripts for model training
│   ├── model_evaluation.py    # Scripts for model evaluation
│   └── utils.py               # Utility functions
├── tests/              # Unit tests and validation scripts
├── requirements.txt    # Python dependencies
├── .gitignore          # Git ignore file
└── README.md           # Project overview and instructions

Installation

To set up the project, follow these steps:

1. Clone the Repository:

   git clone https://github.com/your-username/capsule-vision-2024-challenge.git
   

2. Navigate to the Project Directory:

   cd capsule-vision-2024-challenge
   

3. Install Dependencies:

   Make sure you have pip installed. Then, install the required Python packages:

   pip install -r requirements.txt
   

Data

Training and Validation Data

The training and validation datasets are accessible here.

Testing Data

The test dataset will be released on October 11, 2024.

Usage

Data Preprocessing

Preprocess the data using the following script:

python src/data_preprocessing.py

Model Training

Train the model using the following script:

python src/model_training.py

Model Evaluation

Evaluate the trained model using:

python src/model_evaluation.py

Contact

For any queries, please contact: ask.misahub@gmail.com

License

This project is licensed under the MIT License - see the LICENSE file for details.

---