A personal goal of mine has been to commit more time to committing code. Lately, I've been experimenting with various AI capabilties. Below are some ideas that I have that leverage LLMs, Computer Vision, and Generative AI. I also am making it a point to contribute to an open source community. One that I've been intrigued by and following recently is ComfyUI
Coffee Bot AI
Leveraging an iPhone camera, computer vision, elevenlabs voice, and llama 3.2; I'm cultivating a morning coffee companion that recognizes me, automatically starts my coffee maker, and provides some heartwarmingly fun conversation.
Multimodal Digital Assistant
The inspiration for Coffee Bot AI. This project imagines what an AI avatar could look like and displays multiple emotional facial responses as you interact with it. Leverages llama 3.2 for dictating and natural language processing, elevenlabs for voice, and runway gen3 for animations
Guitar Tab Generator
This project is still in its infancy but I'm exploring the concept of training an LLM on guitar tabs from my favorite artists in hopes of generating inspired new tabs that I can use to practice playing guitar. Leveraging LangChain, Supabase, and a host of LLM providers until I find what I'm looking for
www.comfy.org
ComfyUI is an open source generative AI workflow visualization product that allows users to gain more control over the customization of digital assets that they create. Leveraging open source models like Stable Diffusion or Loras like Flux.1, ComfyUI gives you precise control over the diffusion process.