project_detail.md

🌾 Rice Disease Detection & Chatbot Assistant

An advanced AI-powered web application for detecting rice leaf diseases from images, generating multilingual expert diagnosis, and answering agricultural queries via chatbot—all deployed with Gradio on Hugging Face Spaces.

---

📌 Overview

This app intelligently detects rice leaf diseases using a deep learning model and provides a comprehensive diagnosis report using CrewAI agents. It supports English, Urdu, and Hindi outputs and generates voice notes for farmer-friendly interaction.

---

🎯 Key Features

• ✅ Detect 10 types of rice leaf diseases using a trained CNN model.
• 🧠 Multi-agent diagnosis pipeline using CrewAI + OpenAI API.
• 🌍 Multilingual translation (English, Urdu, Hindi) via Groq API.
• 🔊 Voice note generation using gTTS for easy understanding.
• 🤖 Built-in agriculture-themed chatbot for rice-related queries.

---

🧪 Tech Stack

🎨 Frontend

• Gradio: UI for file upload, language selection, and chatbot.
• Hugging Face Spaces: Deployment platform.

🧠 Backend

• TensorFlow/Keras: Rice disease image classification model.
• CrewAI: Multi-agent system for diagnosis logic.
• OpenAI API: LLM-powered agents (Pathologist, Agronomist, Risk Advisor).
• Groq API: Fast LLM inference and translation.
• LangChain: Tool chaining and agent orchestration.
• gTTS: Text-to-speech audio generation.
• FAISS + SentenceTransformer: Chatbot RAG response system.

---

🧬 Architecture

graph TD
    A[User Uploads Rice Leaf Image] --> B[TensorFlow Disease Prediction]
    B --> C[CrewAI Agent Pipeline]
    C --> D[Diagnosis: Pathologist, Agronomist, Risk Advisor]
    D --> E[Groq Translation to Selected Language]
    E --> F[Text + Voice Output using gTTS]
    A2[User Asks a Rice Question] --> G[Chatbot RAG Engine]
    G --> E

---

🤖 CrewAI Agents

• Pathologist Agent: Explains disease cause and symptoms.
• Agronomist Agent: Suggests treatment and crop care.
• Risk Advisor Agent: Evaluates disease severity and urgency.

---

🌍 Language Support

Language	Text Output	Voice Note
English	✅	✅
Urdu	✅	✅
Hindi	✅	✅

---

🚀 How to Run Locally

1. Clone this repository or download files.
2. Add your API keys in environment variables:
   • GROQ_API_KEY
   • OPENAI_API_KEY
3. Install dependencies:

   pip install -r requirements.txt
   

4. Run the app:

   python app.py
   

---

🏗️ Deployment (Hugging Face Spaces)

1. Create a new Space → Gradio + Python.
2. Upload:
   • app.py
   • model.h5
   • rice_chatbot.py
   • crew_agents.py
   • requirements.txt
3. Add API keys in "Secrets":
   • GROQ_API_KEY
   • OPENAI_API_KEY
4. Click Deploy.

---

📦 File Structure

📁 cropdoctor/
├── app.py                # Main Gradio app
├── model.h5              # Trained rice disease model
├── rice_chatbot.py       # Chatbot component
├── crew_agents.py        # CrewAI agent setup
├── requirements.txt      # Python dependencies

---

🙌 Credits

• 🧑‍💻 Developed by
• 🤖 Powered by Groq API, OpenAI, and CrewAI

---

📜 License

This project is open-source and available under the MIT License.