YOLO models evaluation app using Streamlit

🎥 Click on the image above to see a demo video on YouTube.

Bachelor of Software Engineering
Yoobee College of Creative Innovation, Auckland

Introduction

Welcome to the Comparative Object Detection project! This project, a capstone for the Bachelor of Software Engineering at Yoobee Colleges of Creative Innovations in Auckland, New Zealand 🥝.

Features

Interactive Web Application: The application, built using Streamlit, offers a user-friendly interface. Users can easily upload their data, choose from a variety of YOLO models, and get real-time detection results.
Multiple YOLO Versions: Users have the flexibility to evaluate pre-trained YOLO models from various versions, including v7 and v8.
Comparison: Engage in a detailed comparison between YOLOv7 and YOLOv8 models. Analyze their performance across various metrics, observe differences in detection accuracy and processing speed, and determine the optimal model for different use cases.
User Authentication & Dashboard: With a built-in authentication system, users can securely access the dashboard, view past detection results, and provide feedback.

Installation and Setup

Create a virtual environment

# for Windows
# create 
python -m venv myvenv

activate

myvenv\Scripts\activate

Clone the Repository:

   git clone https://github.com/irinagetman1973/YOLO-Streamlit
   cd YOLO-Streamlit

Install Dependencies:

      pip install -r requirements.txt

Run the Application:

      streamlit run main.py

Download Pre-trained YOLOv8 Detection Weights

Create a directory named weights and create a subdirectory named detection and save the downloaded YOLOv8 & YOLOv7 object detection weights inside this directory. The weight files can be downloaded from the tables below.

Model	size ^(pixels)	mAP^val 50-95	Speed ^{CPU ONNX (ms)}	Speed ^{A100 TensorRT (ms)}	params ^(M)	FLOPs ^(B)
YOLOv8n	640	37.3	80.4	0.99	3.2	8.7
YOLOv8s	640	44.9	128.4	1.20	11.2	28.6
YOLOv8m	640	50.2	234.7	1.83	25.9	78.9
YOLOv8l	640	52.9	375.2	2.39	43.7	165.2
YOLOv8x	640	53.9	479.1	3.53	68.2	257.8

Download Pre-trained YOLOv7 Detection Weights

Model	Test Size	AP^test	AP₅₀^test	AP₇₅^test	batch 1 fps	batch 32 average time
YOLOv7	640	51.4%	69.7%	55.9%	161 fps	2.8 ms
YOLOv7-X	640	53.1%	71.2%	57.8%	114 fps	4.3 ms

YOLOv7-W6	1280	54.9%	72.6%	60.1%	84 fps	7.6 ms
YOLOv7-E6	1280	56.0%	73.5%	61.2%	56 fps	12.3 ms

Getting Started

To begin exploring the capabilities of the YOLO Models Evaluation App, follow these simple steps:

Installation: Follow the installation guide above to set up the application on your local machine.
Model Selection: Choose between YOLOv7 and YOLOv8 models based on your requirements and the nature of your dataset.
Upload Your Data: Easily upload images or video streams to test the models' detection capabilities.
Analyze Results: Utilize the data analysis tools to interpret the models' performance and make informed decisions.
We love your feedback: Join discussions, share feedback, and contribute to the development of the application.

Contributing

We welcome contributions of all kinds - from code improvements and bug fixes to documentation updates. Please feel free to fork the repository, make changes, and submit a pull request. Your contributions will help make this project even better!

License

This project is licensed under the MIT License. Feel free to use, modify, and distribute the code as per the license terms.

Enjoy exploring the world of YOLO models with the YOLO models Comparative project!

🥝 Developed with passion by Irina Getman.

irinagetman1973 / YOLO-Streamlit