EcoSort: Your Waste Sorting Companion 🌱♻️

Managing waste effectively is crucial for the well-being of our environment and communities. EcoSort offers an innovative solution to tackle waste management challenges and promote sustainable practices.

What EcoSort Does

• Trash Sorting Education: Engage in entertaining mini-games to learn about proper waste sorting techniques while having fun and staying motivated.
• Image Recognition: Utilize our user-friendly web application to instantly identify recyclable items by uploading their images.
• Real-time Assistance: Access a chatbot function for on-demand information about trash management techniques.

Why Choose EcoSort?

• Environmental Impact: Contribute to environmental sustainability by adopting recycling and waste reduction practices.
• Community Engagement: Make waste sorting fun and rewarding for individuals and communities with interactive learning tools, including engaging mini-games.
• Technological Innovation: Benefit from cutting-edge image recognition technology to simplify waste sorting.

Key Features

1. Interactive Learning: Engage in entertaining mini-games to learn efficient garbage sorting methods while enjoying the process.
2. Image Recognition: Upload item images to identify recyclable objects rapidly and precisely.
3. Real-time Assistance: Utilize the chatbot function for on-demand information about trash management techniques.

How We Are Building It?

Component	Technology Stack
Frontend	React, JavaScript, CSS
Backend	Python, Django, PyTorch, Transformers
Image Recognition	CNN (Convolutional Neural Network)
Chatbot	Natural Language Processing (NLP)
Mini-Game	Unity

Getting Started

1. Frontend Setup:

   • Navigate to the frontend directory and run it on live server to start the frontend server.

2. Backend Setup:

   • Navigate to the backend directory and set up the transformer and move chatbot.py in huggingface_interface and run python chatbot.py to start the Flask server to run the chatbot.

3. Transformer Setup:

   • Please Follow the following instructions for cloning the transformer

   # Clone the github repository and navigate to the project directory.
     git clone https://github.com/AI4Bharat/IndicTrans2
     cd IndicTrans2
     # Install all the dependencies and requirements associated with the project.
     source install.sh

   • Inside IndicTrans2 clone the tokeniser using the instructions below

   git clone https://github.com/VarunGumma/IndicTransTokenizer
   cd IndicTransTokenizer
   pip install --editable ./

   • Now navigate to example.py in huggingface_interface in IndicTrans2 and please paste this code instead of the existing one

      import torch
      from transformers import AutoModelForSeq2SeqLM
      from IndicTransTokenizer import IndicProcessor, IndicTransTokenizer
   
      tokenizer = IndicTransTokenizer(direction="en-indic")
      ip = IndicProcessor(inference=True)
      model = AutoModelForSeq2SeqLM.from_pretrained("ai4bharat/indictrans2-en-indic-dist-200M", trust_remote_code=True)
   
      sentences = [
          "This is a test sentence.",
          "This is another longer different test sentence.",
          "Please send an SMS to 9876543210 and an email on [email protected] by 15th October, 2023.",
      ]
   
      batch = ip.preprocess_batch(sentences, src_lang="eng_Latn", tgt_lang="hin_Deva")
      batch = tokenizer(batch, src=True, return_tensors="pt")
   
      with torch.inference_mode():
          outputs = model.generate(**batch, num_beams=5, num_return_sequences=1, max_length=256)
   
      outputs = tokenizer.batch_decode(outputs, src=False)
      outputs = ip.postprocess_batch(outputs, lang="hin_Deva")
      print(outputs)
    

Note: We recommend creating a virtual environment with python>=3.7.

Challenges Faced

1. Limited Data Availability: Dealing with diverse data sources required careful consideration to ensure accuracy.
2. User Engagement: Designing interactive and engaging tools, including mini-games, was crucial to encourage participation.

Accomplishments to Be Proud Of

Milestone	Description
Engaging User Interface	Designed a visually appealing interface for enhanced user experience.
Efficient Image Recognition	Developed a reliable image recognition system for seamless waste sorting.
Real-time Assistance	Implemented a chatbot function to provide instant help and guidance.

Lessons Learned

• User-Centric Design: Prioritizing user experience highlighted the importance of intuitive design and functionality.
• Technological Innovation: Leveraging advanced technologies like PyTorch, Transformers, and Unity expanded our capabilities.
• Community Impact: Engaging communities through education and technology, including mini-games, fosters a culture of environmental responsibility.

What's Next for EcoSort?

Next Steps	Description
Expansion of Educational Resources	Develop additional educational content and games to further engage users.
Integration with Local Communities	Collaborate with local authorities and organizations to promote EcoSort adoption.
Continuous Improvement	Gather user feedback to refine features and enhance usability.