Inspiration

The inception of FLOW stemmed from a personal affinity for meditation, which took a purposeful turn upon encountering a friend with Aphantasia during high school. The realization that individuals with Aphantasia miss out on the enriching experience of meditation spurred the creation of FLOW, an app crafted to bridge this gap.

What it does

FLOW is an immersive platform allowing users to tailor a VR scene and meditation guidance to their liking using natural language. It retains a history of past VR sessions, enabling users to revisit their customized meditative realms.

How we built it

Constructed using Flutter with Firebase handling the backend operations, FLOW's VR facet is brought to life through Aframe.js and Three.js. The VR scenes are hosted on a website and are integrated into the app via webview. Utilizing large language models alongside Neural2 text-to-speech technology, FLOW offers a human-like, fully customizable meditation guidance. A pipeline from the Diffusion model translates user inputs into panoramic images, enriching the VR experience.

Challenges we ran into

Transitioning a customized PyTorch model through pipelines to a TensorFlow Lite model for integration into our Flutter app posed a significant challenge.

Accomplishments that we're proud of

The completion of an expansive project like FLOW within the stringent timeline of a hackathon is an achievement that brings immense pride.

What we learned

The journey imparted valuable insights into overcoming technical hurdles and innovatively merging different technologies to create a holistic solution.

What's next for FLOW

The roadmap ahead for FLOW involves forging collaborations and partnerships with Aphantasia and mental health organizations worldwide, aiming to unveil the potential of FLOW to those who stand to benefit the most from it.

Built With

Share this project:

Updates