Symbiosis Project

Designing an adaptive room for captivating the collective consciousness from internal states.

Research paper

Longevity, AI, and Cognitive Research Hackathon at MIT Media lab by ekkolapto and Augmentation lab

For demo purposes as a workflow showcasing, multimodal language models (MMLMs) were implemented via API requests.

Features shown:

• Video capturing with opencv
• Image analysis with gemini-1.5-flash, delivers a description for the image generator.
• Image generation with dall-e-2
• Image display with PIL
• Binaural beats generation with sounddevice

Recorded demos (no audio capturing):

Screencast.from.2024-10-27.14-56-43.webm

Screencast.from.2024-10-27.14-52-53.webm

Running instructions

Test system: Laptop with GTX 1650, on Ubuntu 24.04.1 with Python 3.12

• Install Tkinter from apt: sudo apt-get install python3-tk
• Install Portaudio from apt: sudo apt-get install portaudio19-dev
• Clone the repository
• Create a virtual environment: python -m venv env
• Activate virtual env: source env\bin\activate
• Install libraries: pip install -r requirements.txt
• Create .env file, with API keys (OPENAI_API_KEY,GOOGLE_API_KEY)
• Run: python3 hack-demo.py