---
title: Accident_Detection_App
emoji: 😏
colorFrom: blue
colorTo: red
sdk: streamlit
app_file: accident_app.py
pinned: false
---


# Accident Detection Model

This application showcases the capabilities of our Accident Detection Model, a pivotal component of our research project focused on Accident Detection within Smart City Transportation frameworks.

## Overview

The application empowers users to view a selection of sample accident videos and upload a new video to test the model. Our model is adept at detecting accidents in both trimmed and untrimmed video formats.

## Table of Contents

- [Installation](#installation)
- [Usage](#usage)
- [Features](#features)
- [Contribution](#contribution)
- [License](#license)
- [Acknowledgments](#acknowledgments)

## Installation

1. **Clone the repository:**
    ```bash
    git clone [(https://github.com/adewopova/Accident_detection_SM_City/)]
    ```

2. **Navigate to the directory:**
    ```bash
    cd path_to_diretory
    ```

3. **Install the required dependencies:**
    ```bash
    pip install -r requirements.txt
    ```

4. **Launch the Streamlit app:**
    ```bash
    streamlit run app.py
    ```

## Usage

With the app up and running:

- Opt between trimmed and untrimmed video variants.
- Pick a sample video from the provided list or upload a video of your choice.
- The model will analyze the video and superimpose accident likelihood indicators.

## Features

- **Sample Videos**: Preloaded sample videos for immediate testing.
- **Accident Prediction**: The core functionality that exhibits the probability of an accident occurrence within the selected video.
- **User-friendly Interface**: Crafted using Streamlit, ensuring a seamless and intuitive user experience.

## Contribution

Your contributions can make a difference! Kindly consult the contribution guidelines prior to submitting any changes.

## License

This project is protected under the MIT License. For more details, please refer to the `LICENSE.md` file.

## Acknowledgments

A heartfelt appreciation to our dedicated research team members: Victor Adewopo and Nelly Elsayed. 

[https://arxiv.org/pdf/2310.10038.pdf](#)