Object Detection w/ YOLO v8

This as an implementation of Ultralytics YOLO v8 for object detection.

Introduction

YOLO v8 is state of the art model for computer vision tasks from Ultralytics. In this project, I have implemented it for object detection task.

Features

1. Real time object detection via Webcam
2. Object detection in an uploaded video file

Technologies Used

1. Ultralytics YOLO v8 (Model)
2. OpenCV (Video Processing)
3. Roboflow Supervision (Video Processing)
4. Streamlit (UI)

Project Structure

├─── models/
│     └─── __init__.py
├─── outputs/
├─── research/
│     └─── trials.ipynb
├─── sources/
├─── utils/
│     └─── __init__.py
├─── variables/
│     └─── __init__.py
├─── app.py
├─── template.py
├─── README.md
├─── requirements.txt
├─── LICENSE
└─── .gitignore

Dependencies

streamlit~=1.26.0
ultralytics~=8.1.1
supervision~=0.17.1
opencv-python~=4.9.0.80
pathlib~=1.0.1
pillow~=9.4.0
tqdm~=4.65.0
pyyaml~=6.0

How to Install and Run

1. Clone the repository.

$ git clone https://github/repository/link

2. Get into the Project Directory

$ cd Object-Detection-w-YOLOv8

3. Install the required dependencies.

$ pip install -r requirements.txt

4. Run the app.

$ streamlit run app.py

5. Choose the mode and model.
6. Upload a video if Video File mode is selected.
7. Click the Start Detecting button.

Use Cases

• It can be used to count cars entered and left a parking lot.
• To count live stock.
• Counting people that entered and left a shop.
• Other than counting, it can be used to detect any motion in a frame.
• And much more.

Contributing

Contributions are welcomed and appreciated.

License

This project is released under 'MIT License'.

Conclusion

In conclusion, this project leverages the power of Ultralytics YOLO v8 for real-time object detection. With features like webcam-based detection and processing of uploaded video files, it offers a versatile solution for various applications.