AniMix 🌿 — Touch Grass, Photograph Real Animals & Fuse Them into AI Monster Cards!

Description and use

AniMix lets users photograph wildlife, classify animals locally in the browser, and fuse them into custom monster cards, all offline and privacy‑preserving.

Problem

Users want to identify real animals in the wild and create custom monster cards without relying on cloud APIs or compromising privacy.

Outcome

AniMix demonstrates that offline, client‑side AI can enable interactive, privacy‑preserving experiences in remote environments, achieving zero API cost and unlimited scalability.

Workflow

• Explore outdoors and capture photos with a smartphone or laptop camera.
• Run the in‑browser MobileViT‑small model via Transformers.js to classify the animal species instantly.
• Select two captured animals from the deck and fuse them to generate a custom monster card with procedurally calculated stats and element affinities.
• Export the card as a high‑DPI image or share it using the Web Share API.
• Build a deck or share cards with friends.

Evidence

AniMix runs an open-weight vision classification model directly inside your browser to identify the animal species instantaneously.

Source

The model runs entirely in-browser via WebAssembly and ONNX Runtime, performing vision classification without any backend API call.

Source

When hiking in remote parks or nature trails with zero cell signal, traditional cloud vision APIs fail completely.

Source

Outdoor photos of surroundings stay on the user's device. No images are ever uploaded to third-party servers.

Source