This is a submission for the Google AI Studio Multimodal Challenge
What I Built
I built Chroma Canvas, your personal AI-powered art studio.
At its heart, Chroma Canvas is designed to solve a universal problem: the blank canvas. It’s for anyone who has ever felt a spark of creativity but didn't know where to start, or for those who see the world in a more artistic light.
It’s more than just an app; it's a creative partner that blends the real world with your wildest imagination.
Chroma Canvas has two distinct modes:
🎨 *Paint a Place*
This mode turns the entire globe into your canvas. You can input any address, and using Google Maps satellite imagery, Chroma Canvas transforms that real-world location into a beautiful piece of art. You're not just getting a filtered photo; you're getting a unique creation in styles like Watercolor, Oil Painting, Sketch, Steampunk, and more.
✨ *Imagine a Scene*
This is where your imagination runs wild. Powered by Imagen, this text-to-image mode lets you describe any scene, concept, or dream, and brings it to life in stunning detail. From "a cyberpunk city in the rain" to "a serene bioluminescent forest," the only limit is what you can type.
But it doesn't stop there! I've also built in a suite of AI-powered tools to make the experience truly interactive:
- 🪄 Magic Edit: Edit your creations with simple text prompts.
- 🧐 Vincent, the AI Critic: Get a witty critique of your art.
- 🎨 Palette Prodigy: Extract the dominant color palette from your masterpiece.
- 💡 AI Location Genius: Learn fascinating facts about the places you paint.
Demo
You can try Chroma Canvas live right here:
Link to Deployed Applet
Here’s a glimpse of what you can do:
From Satellite to Seurat: Painting a Place
Here we have the stunning Sajjangarh Monsoon Palace in Udaipur. On the left is the satellite view from Google Maps, and on the right is the magical watercolor painting generated by the AI.
Imagination Unleashed: Creating a Scene
I prompted the "Imagine a Scene" mode with: "A majestic bioluminescent jellyfish floating in a nebula, cinematic, detailed" and got these incredible results.
The Magic of Nano Banana: Editing an Artwork
Here's a short clip showcasing the Magic Edit feature. I took the original artwork and simply asked it to "add a dragon flying in the sky". The result is seamless and inspiring.
(A video placeholder showing the original artwork, a user typing the edit prompt, and the artwork regenerating with a majestic dragon soaring through the clouds.)
How I Used Google AI Studio
Google AI Studio was the creative engine behind every core feature of Chroma Canvas. I leveraged a suite of powerful multimodal models to build a seamless and feature-rich experience.
gemini-2.5-flash-image-preview(Nano Banana): This was the star of the show for the Paint a Place and Magic Edit features. Its ability to take an input image (like a map screenshot) alongside a text prompt is the core of the app's "artistic transformation" magic.imagen-4.0-generate-001: This model is the powerhouse behind the Imagine a Scene mode, translating pure text descriptions into high-quality, imaginative visuals.-
gemini-2.5-flash: This versatile model was the workhorse for all the "intelligent" text-based and analytical features. I used it for:- Generating witty art critiques.
- Providing interesting descriptions of real-world locations.
- Suggesting famous landmarks to inspire users.
- And, most impressively, analyzing an image to extract a color palette and return it in a structured JSON format, thanks to the
responseSchemafeature.
Multimodal Features
The true power of Chroma Canvas comes from its deep integration of multimodal AI, creating an experience that's more than the sum of its parts.
The app creates a constant, fluid dialogue between images and text, which I believe is the future of creative tooling.
Here’s how the multimodal features enhance the user experience:
-
Image + Text to Image Transformation:
- What it is: In "Paint a Place," the app takes a satellite image and a text prompt (the art style) to generate a completely new image.
- Why it's great: It’s a magical experience. It transforms a literal, data-rich map view into an emotional, artistic interpretation. It empowers users to see their own home, or any place on Earth, through the eyes of an artist.
-
Conversational Image Editing:
- What it is: The "Magic Edit" feature allows users to refine a generated image using simple text commands.
- Why it's great: This turns the creative process from a one-shot command into an interactive conversation. Users don't have to be prompt engineering experts; they can iteratively build on their creation, making changes as naturally as if they were speaking to a fellow artist.
-
Image Analysis for Structured Data:
- What it is: The "Palette Prodigy" feature takes an image and asks the model to output structured JSON data representing the color palette.
- Why it's great: This is a fantastic example of multimodal understanding. The AI isn't just "seeing" the image; it's analyzing it and converting its visual properties into useful, tangible data. This is incredibly valuable for artists and designers looking for inspiration.
-
Image Analysis for Creative Text:
- What it is: "Vincent, the AI Critic" looks at a generated image and writes a short, witty critique in text.
- Why it's great: This adds a delightful, interactive layer that brings the user's creation to life. It makes the AI feel less like a tool and more like a quirky, creative partner, making the experience more memorable and fun.
By weaving these different multimodal capabilities together, Chroma Canvas aims to be a truly intuitive and powerful partner for anyone looking to unlock their creativity.
Top comments (1)
Very innovative !