Accident Anticipation Project - Setup and Guidelines

This repo is for the rough works for our major project.

Find the reference papers used in this project here

Project Setup

1. Clone the Repository

Start by cloning this repository and setting up the Python environment.

2. Create a Conda Environment

Ensure you have Conda installed. Create a new environment with Python 3.10:

conda create -n accident-anticipation python=3.10

conda activate accident-anticipation

If you're using PyCharm, select this environment as the project's interpreter.

3. Install PyTorch and Dependencies

For systems with CUDA, verify your CUDA version and find the corresponding PyTorch installation command on the PyTorch Get Started page.

For example, with CUDA 12.4, use:

conda install pytorch torchvision torchaudio pytorch-cuda=12.4 -c pytorch -c nvidia

Check your system compatability before installation

---

Guidelines

Avoid Pushing Large Binary Files

Do not push the following to the repository:

• Dataset images
• Saved model files
• Other large binary files

To prevent accidental uploads, ensure these files are listed in the .gitignore file.

Begin writing code in the src directory.