My Project

Threat Detection in Videos using Faster R-CNN 🎥🚨

Overview

This repository provides a Threat Detection solution using the Faster R-CNN (Region-based Convolutional Neural Network) model for analyzing videos. The model is trained to detect threats within video frames, making it a valuable tool for security and surveillance applications.

📋 Table of Contents

1. Introduction
2. Setup
3. Training
4. Testing
5. Validation
6. Usage
7. Contributing
8. License

🚀 Introduction

The Faster R-CNN model is employed to identify threats within video frames. This solution allows for efficient threat detection and is customizable based on your dataset.

⚙️ Setup

To get started, follow these steps:

• Clone this repository to your local machine.
• Install the required dependencies by running pip install -r requirements.txt.
• Ensure that your dataset is organized with CSV annotations and image directories.

🚂 Training

To train the model, use the following command:

python train.py -cp <path_to_csv> -ip <path_to_images> -mp <main_directory_path> -simg <sample_image_count> -e <epochs> -sm <save_model_flag>

• -cp or --csv_path: Path to the CSV file containing annotations.
• -ip or --image_path: Path to the directory containing images.
• -mp or --main_directory_path: Path to the main directory.
• -simg or --sample_image: Number of total data samples.
• -e or --epochs: Number of training epochs.
• -sm or --save_model: Specify True if you want to save the trained model.

🧪 Testing

To test the model, use the following command:

python test.py -fp <file_path_to_test_image> --test_model True

• -fp or --file_path: Path to the image for testing.
• --test_model: Specify True to enable testing mode.

✔️ Validation

Validation is an integral part of the training process, ensuring the model's generalization. During training, a portion of the dataset is used for validation. Adjust the parameters accordingly to achieve optimal performance.

🛠️ Usage:

[