/ecotale1

Primary LanguageTypeScriptMIT LicenseMIT

Ecotales.ai

YoloV8, GPT4.0, DallE, NextJS

Project Logo

Welcome to the Ecotales YOLOv8 Tech Stack repository! This project showcases the seamless integration of technologies like TypeScript (TSX), Tailwind CSS, and OpenAI's YOLOv8 in a Next.js application deployed on Vercel. With this repository, you'll be able to kickstart your journey into building intelligent image analysis applications within an eco-friendly narrative.

Table of Contents

Introduction

Ecotales YOLOv8 Tech Stack demonstrates the power of combining Next.js with TypeScript (TSX), Tailwind CSS, and OpenAI's YOLOv8 model. This stack allows you to create web applications that analyze images using state-of-the-art object detection capabilities while maintaining a beautiful and responsive user interface.

Features

  • Modern Tech Stack: Build your project using Next.js and TypeScript, ensuring type safety and scalability.
  • Sleek UI with Tailwind CSS: Design your UI using the utility-first CSS framework, Tailwind CSS, for quick and responsive styling.
  • Intelligent Object Detection: Leverage OpenAI's YOLOv8 model to perform accurate and efficient object detection on uploaded images.
  • User-friendly Interface: Provide an intuitive interface for users to upload images, view detections, and interact with the results.
  • Easy Deployment: Deploy your application effortlessly using Vercel's seamless deployment pipeline.

Why Did We Come Up With This Idea?

Climate change - a crucial worldwide issue that causes rising sea levels and extreme weather patterns. Singapore, a low-lying island state, is particularly vulnerable to the effects of climate change. To tackle this challenge, Singapore has implemented the SG Green Plan 2030. Launched in February 2021, the Singapore Green Plan 2030 seeks to galvanise a whole-of-nation movement and advance Singapore’s national agenda on sustainable development. As a small and resource-scarce city-state, sustainable development is only achieved by collective efforts of the whole nation. That’s why we want to inculcate the notion of going green to young children, which is the future generation who plays an important role in fighting climate change.

Amid the vibrant streets of Singapore, I uncovered a jarring truth: 7.7 million tonnes of waste discarded in 2018, filling 15,000 Olympic pools. Yet, Singapore's SG Green Plan 2030 ignited hope—a carbon-neutral vision. Casting aside statistical shock, I saw potential within children, the torchbearers of transformation.

Thus, "EcoTales" was born. An educational website to educate young children on how to classify different types of wastes, thus being able to dispose and recycle them properly. Our theme is Circular Economy and Sustainable Living. The website makes use of Machine Learning, Computer Vision and Natural Language Processing to detect the waste object, after that, it will prompt a question to ask the user and provide information about the waste.

Our Objectives

  • Progress towards SG Green Plan 2030 (Pillar 3: Sustainable Living)
  • Reduce waste and increase recycling rate
  • Raise awareness of young children about the importance of waste management
  • Educate young children on how to classify different types of waste

What It Does

  • Webcam will scan the waste material
  • Capture image of the waste
  • Compare the image with the pre-trained model using Machine Learning
  • Sending the command to ChatGPT 3.5
  • Prompt question to ask the users the type of waste
  • Convert image of the waste into cartoon gif
  • Suggest fun facts about the waste.

Challenges and Accomplishments

Insufficient computational resources posed a challenge. Absent a GPU for timely model training to enhance accuracy, we pivoted. Instead, we harnessed a finely-tuned model, crafted on a targeted waste classification dataset.

Owing to the constraints posed by both funding limitations and time constraints, our ability to extensively test the API call is restricted. Regrettably, the scope of our experimentation is confined to a limited number of attempts. These constraints, while challenging, have compelled us to strategize meticulously and make the most out of each testing iteration. With a limited number of trials at our disposal, we must prioritize our testing parameters, meticulously select our data inputs, and implement systematic variations in our testing process.

Future Improvements

  • Scale up the system to reach out to more users
  • Train the object detection model to be more accurate when detecting an object
  • Extend the website with multimodels, such as the use of AI assistant (audio and image) while prompting the questions to reach as younger target audience (kindergartens)

Getting Started

To get started with Ecotales YOLOv8 Tech Stack, follow these steps:

Installation

  1. Clone this repository to your local machine:
git clone https://github.com/pham0084/ecotale1.git
cd ecotale1

  1. Open your browser and navigate to http://localhost:3000 to see the application in action.

  2. Upload an image and experience the magic of YOLOv8-powered object detection.

Usage

Run the development server:

npm run dev

Contributing

We welcome contributions from the community! To contribute to Ecotales YOLOv8 Tech Stack, follow these steps:

  1. Fork this repository to your GitHub account.

  2. Create a new branch with a descriptive name for your feature or bug fix.

  3. Make your changes and commit them with clear and concise commit messages.

  4. Push your changes to your forked repository.

  5. Create a pull request (PR) to the main branch of this repository.

  6. Our team will review your PR, provide feedback, and work with you to get your changes merged.

License

This project is licensed under the MIT License.

Feel free to explore, experiment, and contribute to Ecotales YOLOv8 Tech Stack. Create engaging applications that make a positive impact on the environment and technology landscape.

For questions or assistance, contact us at ecotales.ai

Happy coding!