zkBuddy

zkBuddy is a private AI chatbot that integrates emotion recognition and zero-knowledge proofs (zkML) into a single Streamlit application. It uses Gemini's Generative AI to generate chat responses, YOLO to detect user emotions from images, and a Rust module (built with PyO3 and maturin) to compile cryptographic circuits and simulate GKR-based proofs. This approach ensures that sensitive computations are verified in zero-knowledge without exposing the underlying data.

Features

• Chatbot Interface: Powered by Gemini Generative AI.
• Emotion Recognition: Detects user emotions using a YOLO model.
• Zero-Knowledge Proofs: Simulates circuit compilation and zkML proof generation with Polyhedra’s Expander Compiler Collection.
• Streamlit UI: An interactive web interface for easy interaction.

Requirements

• Python 3.12 (ensure your virtual environment uses this version)
• Rust (latest stable or nightly)
• maturin for building the Python extension from Rust
• Streamlit (pip install streamlit)
• Google Generative AI (with a valid API key)
• Ultralyitcs YOLO (with a pre-trained model file best.pt)
• Polyhedra’s Expander Compiler Collection (declared as a git dependency in Cargo.toml)

Installation

1. Set Up the Virtual Environment

1. Clone the repository and navigate into it:

   git clone https://github.com/ardor05/zkBuddy.git
   cd zkBuddy

2. Create and activate a virtual environment:

   Windows (PowerShell):

   python -m venv .venv
   .\.venv\Scripts\Activate

   macOS/Linux:

   python -m venv .venv
   source .venv/bin/activate

3. Install Python dependencies:

   pip install streamlit google-generativeai ultralytics

2. Build the Rust Module

1. Ensure Rust is installed via rustup.

2. Install maturin:

   pip install maturin

3. Build and install the Rust module into your virtual environment:

   maturin develop -b pyo3

   Note: The Rust module is built with PyO3 and installed as a Python package (e.g., zk_buddy). Check pip list to confirm installation.

Usage

Running the Streamlit App

With your virtual environment activated, run:

python -m streamlit run .\main.py

• Chat Interface: Type a message to receive a response from Gemini.
• Emotion Detection: Upload an image using the sidebar to detect emotions.
• zkML Proof Logging: The Rust module compiles the cryptographic circuit and simulates a GKR proof. The proof details are printed to the terminal (and are hidden from the UI).

Project Structure

zkBuddy/
├── Cargo.toml           # Rust project configuration
├── pyproject.toml       # Maturin build configuration
├── README.md            # This file
├── main.py              # Streamlit application
├── src/
│   └── lib.rs           # Rust source code for zkBuddy
├── .venv/               # Python virtual environment (created locally)
└── best.pt              # YOLO model for emotion detection

Key Technologies

• Expander Compiler Collection by Polyhedra:
  The heart of zkBuddy is built on Polyhedra’s Expander Compiler Collection. The Rust code in src/lib.rs leverages this technology to define and compile cryptographic circuits (for example, a simple circuit that asserts x equals y). It further simulates GKR proof generation and exposes a function (run_zkbuddy) to Python via PyO3, enabling zero-knowledge proof capabilities without revealing sensitive data.

• Gemini AI:
  zkBuddy integrates Gemini’s Generative AI to produce dynamic and contextually rich chat responses. This technology drives the conversational aspect of the chatbot, ensuring that users receive intelligent and relevant replies.

• YOLO (You Only Look Once):
  For emotion recognition, zkBuddy employs a YOLO model to analyze uploaded images. This enables the system to detect and interpret user emotions, which are then used to adjust the chatbot’s responses accordingly.

• Python & Streamlit:
  The Python code in main.py provides a user-friendly Streamlit-based interface that brings together the power of Gemini AI, YOLO for emotion detection, and the underlying Rust-based cryptographic proof system.

Contributing

Contributions are welcome! Please open an issue or submit a pull request on GitHub.

Disclaimer

This project have been developed for Explore Expander Bootcamp by Polyhedra & organized by Encode Club. Thank you to all people that contributing to zkBuddy.