Hues

Roshni Rane — Mon, 12 Oct 2026 06:59:17 +0000

This is a submission for the Hacktoberfest Open-Source AI Challenge Week 1: Touch Grass

Hues - An atlas of concealed colour palettes.

🌿 What I Built

We walk past thousands of colours every day without really seeing them.The muted green of an old gate. The dusty terracotta of a weathered wall. The soft yellow of sunlight falling on a leaf.

Hues is a pocket colour archivist designed to help you notice them.

It turns an ordinary outdoor observation into a small act of discovery: step outside, find something that catches your eye, capture its colour, and preserve it in a growing personal atlas of palettes.

Rather than treating colour as another asset to search for on a screen, Hues begins with the world around you.

The idea is simple: the everyday has a palette of its own. You just have to look.

Hues is for curious observers, designers, artists, and anyone who wants to find a little more wonder in familiar surroundings.

🎨 Demo

🎥 Video Demo: Demo Video link

drive.google.com

The best way to experience Hues is to take it outside. Find a colour in your surroundings, capture it, and start building your own collection.

💻 Code

🔗 GitHub Repository: https://github.com/rrane1307/Hues

The project is being developed with an emphasis on keeping the experience approachable, the interface playful, and the act of collecting colours more meaningful than simply saving hexadecimal codes.

🛠️ How I Built It

I approached Hues as a small digital companion to a physical-world activity, rather than another elaborate productivity tool.
The design centres on the act of collecting: discovering a colour, recording it, and giving it a place in a personal archive.

The application itself is built with
Technology stack :

Frontend: React 19 with JavaScript (ES modules)

Styling: Custom CSS

Build tool: Vite 8 with the React plugin

Browser storage: IndexedDB for specimens and treasures; localStorage for onboarding state

Color sampling: Browser camera APIs and the HTML Canvas API

Optional AI naming: Google DeepMind’s Gemma 2 2B Instruct through the Hugging Face Inference API when a user provides a token; otherwise, local deterministic heuristics generate suggestions

Linting and package management: Oxlint and npm

One design decision guided the process throughout: the interface should invite exploration without becoming the main attraction. The colours are the specimens; the interface exists to help you notice, preserve, and revisit them.

🌍 Why Does Open Innovation Matter?

A project about noticing the world shouldn't depend on an opaque system to tell you what to see.

Open innovation makes it possible to experiment with different models, understand how the underlying tools work, and adapt the experience without being locked into a single proprietary ecosystem.

For Hues, this matters because the project is intentionally small and personal. Its value lies in the experience it creates, not in hiding complexity behind a black box.

Open tools also make the project easier for other developers to inspect, extend, and contribute to. Someone could experiment with new ways to organise colour collections, improve colour extraction, or explore different approaches to building a personal visual archive.

The goal isn't to add AI everywhere. It's to use open technology where it genuinely helps, while keeping the experience grounded in the world outside the screen.

Prize Categories

Hacktoberfest Open-Source AI Challenge Week 1: Touch Grass (Primary Track)
Google Gemma / Open-Weights AI Track