[Project Addition]: Human Detection on Railway Tracks for Suicide Prevention

[vishuhere]

Deep Learning Simplified Repository (Proposing new issue)

:red_circle: Project Title : Suicide Prevention on Railway Tracks
:red_circle: Aim : I propose implementing a feature within the existing railway safety systems to detect individuals on railway tracks using an advanced computer vision algorithm. This feature would analyze live camera feeds along railway tracks and identify human presence, alerting railway authorities or relevant emergency services promptly.
:red_circle: Dataset : To address person detection on railway tracks, we can either propose CNN from scratch or leverage a transfer learning approach with pre-trained models like YOLOv5 or ResNet50, employing OpenCV for robust implementation. Our approach involves training on a curated dataset specific to this project, fine-tuning models for accurate detection. Additionally, integrating an alert mechanism ensures timely notifications to authorities upon detecting individuals on tracks.

---

📍 Follow the Guidelines to Contribute in the Project :

• You need to create a separate folder named as the Project Title.
• Inside that folder, there will be four main components.
  • Images - To store the required images.
  • Dataset - To store the dataset or, information/source about the dataset.
  • Model - To store the machine learning model you've created using the dataset.
  • requirements.txt - This file will contain the required packages/libraries to run the project in other machines.
• Inside the Model folder, the README.md file must be filled up properly, with proper visualizations and conclusions.

---

:red_circle::yellow_circle: Points to Note :

• The issues will be assigned on a first come first serve basis, 1 Issue == 1 PR.
• "Issue Title" and "PR Title should be the same. Include issue number along with it.
• Follow Contributing Guidelines & Code of Conduct before start Contributing.

---

:white_check_mark: To be Mentioned while taking the issue :

• Full name : Vishwajeet Gehlot
• GitHub Profile Link : Github.com/vishuhere
• Email ID : vishwajeetgehlot3@gmail.com
• Participant ID (if applicable):
• Approach for this Project : Mentioned
• What is your participant role? GSsoC Contributor

---

Happy Contributing 🚀

All the best. Enjoy your open source journey ahead. 😎

[github-actions[bot]]

Thank you for creating this issue! We'll look into it as soon as possible. Your contributions are highly appreciated! 😊

[abhisheks008]

You need to implement at least 3 models for this project and find out the best fitted one by comparing the accuracy scores of the models implemented. Please mention the models you are planning to implement here for this problem statement.

@vishuhere

[vishuhere]

Model: YOLOv8

Had worked on this project earlier and YOLOv8 is the best fitted model So far.

It has one of the fastest inference speed, and could be accelerated using GPU.

[abhisheks008]

Model: YOLOv8

Had worked on this project earlier and YOLOv8 is the best fitted model So far.

It has one of the fastest inference speed, and could be accelerated using GPU.

What will be the other two models for this project?

[vishuhere]

Sure! I apologize for misunderstanding what you were trying to convey me earlier.

-Here are the top 3 models we can implement and compare their performance:

1) YOLOv8 (You Only Look Once, Version 8):

The latest iteration in the YOLO family, renowned for its real-time object detection capabilities, offering a balance of high accuracy and speed.

2) Faster R-CNN (Region-based Convolutional Neural Networks):

A two-stage object detection model known for its precision. It first proposes regions where objects might be located and then classifies those regions.

3) EfficientDet:

A state-of-the-art object detection model that achieves a balance between accuracy and efficiency through compound scaling, optimizing both the backbone network and the feature network.

[abhisheks008]

Assigned @vishuhere