This is a submission for the Hacktoberfest Open-Source AI Challenge Week 1: Touch Grass
🌿 WildSphere AI — Hear Nature. See Life. Explore the Planet.
What if AI could do more than keep us staring at screens? What if it could help us look up, step outside, and discover the living world around us?
That question inspired WildSphere AI, an AI-powered biodiversity explorer that connects open-weight AI, wildlife discovery, and real-world nature exploration.
🌎 What I Built
WildSphere AI is designed to help people discover animals, birds, and plants through photographs, wildlife sounds, and an interactive world map.
The idea is simple: use AI to spark curiosity about nature, then encourage people to explore beyond their screens.
Here is what the project brings together:
- 📸 Visual Species Discovery: Upload a photograph to explore a possible animal, bird, or plant identification, along with identifying characteristics and educational facts.
- 🐦 Soundscape Explorer: Analyze bird-call recordings to discover possible species and learn more about the sounds of nature.
- 🗺️ Interactive Biodiversity Map: Explore countries around the world and discover representative animals, birds, and plants associated with their biodiversity.
- 🌱 WildQuest: Get small outdoor challenges, such as observing different leaf shapes, listening for bird calls, or noticing pollinators from a safe distance.
- 📖 Nature Journal: Keep a personal collection of discoveries and reflections.
The project is intended for curious learners, nature enthusiasts, students, and anyone who wants to understand biodiversity in a more interactive way.
Rather than making the screen the destination, WildSphere AI aims to make it the starting point for an outdoor adventure.
🎬 Demo
Demo video: https://drive.google.com/drive/folders/1H-noCxn5yVMgO0Whp69wObw3Tz7wyUl2?usp=sharing
The demo will showcase the discovery workflow, country-level biodiversity exploration, and how a digital discovery can become a real-world nature mission.
💻 Code
GitHub repository: https://github.com/VanshikaTyagi28131/WildSphereAI
The project brings together a web interface, backend AI integrations, biodiversity information, and an interactive exploration experience.
The repository will document the setup instructions, model configuration, data sources, and current limitations so others can explore and build on the project.
🛠️ How I Built It
WildSphere AI is designed around a modular architecture, with different tools handling the tasks they are best suited for.
- Frontend: React, TypeScript, Vite, and Tailwind CSS for a responsive, nature-inspired interface.
- Backend: Node.js and Express for handling application requests and securely connecting external services.
- Open-weight AI: Google's Gemma model for image-based explanations, species context, and educational nature missions.
- Sound identification: A specialist bird-audio classification integration, such as BirdNET, for analyzing wildlife recordings.
- Biodiversity data: Sources such as iNaturalist for species observations and relevant biodiversity information.
- Nature exploration: An interactive country map, species profiles, outdoor challenges, and a personal nature journal.
One important architectural decision is to separate species identification from educational explanation. A specialist model is better suited to identifying bird calls, while a vision-capable model can help interpret photographs and explain the observations in accessible language.
Biodiversity facts and species information should be grounded in external sources wherever possible, rather than relying exclusively on AI-generated claims.
🔓 Why Does Open Innovation Matter?
Biodiversity is too fascinating to be limited to expensive tools or closed AI ecosystems.
Open-weight models such as Gemma create opportunities for developers to experiment with model behavior, build specialized applications, and explore self-hosted inference when hardware and licensing permit.
For WildSphere AI, open innovation matters in three ways:
1. More control over the technology
A modular architecture makes it easier to experiment with different compatible models and specialized classifiers instead of tying the entire application to one AI provider.
2. Opportunities for privacy-conscious and offline experiences
Open-weight models create a path toward local inference, where supported models can run on suitable devices without sending every photograph to a remote AI service. This is an important future direction for nature exploration in places with limited connectivity.
3. Making biodiversity education more adaptable
Developers and contributors can extend the experience with new species datasets, regional nature guides, additional languages, and specialized models for different wildlife sounds.
For me, open innovation is not just about using an accessible model. It is about having the flexibility to understand, adapt, and improve
Top comments (0)