/AI-ATL-2024

Primary LanguagePython

InstaRizz

InstaRizz is a real-time, wearable-powered social assistant that captures live video from Ray-Ban smart glasses, identifies individuals, and generates personalized bios and pickup lines in real-time through a custom AI pipeline. This project was built for a hackathon to explore the intersection of AI, social interaction, and wearable technology.

🚀 Features

  • Live Video Streaming: Streams live video from Ray-Ban smart glasses directly to Instagram Live.
  • Real-Time Facial Recognition: Captures frames from the live stream and identifies individuals using OpenCV.
  • Custom Identity Search: Runs a custom classifier on our database to identify individuals based on their facial data.
  • AI-Generated Insights: Uses Magic Loops API and GPT-4 to create a short bio and generate three pickup lines, displaying them instantly for the user.

🛠 Tech Stack

  • Ray-Ban Smart Glasses: Wearable hardware for capturing video and live streaming.
  • OpenCV: Used for real-time facial recognition on captured video frames.
  • Magic Loops: Hosts a custom API pipeline for performing perplexity search to gather data on individuals and integrates with GPT-4.
  • GPT-4 API: Generates bios and pickup lines based on individual data retrieved by Magic Loops.
  • Streamlit: Provides a quick and interactive UI for showcasing InstaRizz's functionalities, including real-time data display.

📦 Setup

  1. Clone the Repository:

    git clone https://github.com/username/InstaRizz.git
cd InstaRizz

  2. Install Dependencies:

    • Python 3.8+
    • Install required packages: 
      pip install -r requirements.txt

  3. Configure API Keys:

    • Set up environment variables for Magic Loops, GPT-4, and any other required API keys: 
      export MAGIC_LOOPS_API_KEY='your_magic_loops_api_key'
export GPT4_API_KEY='your_gpt4_api_key'

  4. Run the Application:

    streamlit run app.py

🚧 Challenges & Solutions

  • Latency in Real-Time Processing: We optimized snapshot intervals and OpenCV processing to handle data faster.
  • Privacy Considerations: Ensured ethical use by focusing on public data and limiting private info retrieval.

🤔 Lessons Learned

InstaRizz taught us the importance of optimizing real-time data, managing privacy in AI-driven applications, and finding the right balance between speed and accuracy for social tech.

📅 Future Plans

  • Enhanced Privacy Controls: Add more filters for ethical usage and data transparency.
  • Partnership with Instagram: Explore a dedicated InstaRizz feed for richer user interactions.

📜 License

MIT License

InstaRizz is a playful step into AI-driven social engagement, blending real-time tech with personalization. Contributions and feedback are welcome!