Madras Field Notes — A Fully Offline, In-Browser Field Guide for Chennai's Birds and Plants

Description and use

Madras Field Notes is an offline PWA that identifies Chennai’s birds and plants using a browser‑based CLIP classifier and a local Qwen LLM, logs observations locally, and supports English and Tamil.

Problem

Users in Chennai need a portable, offline tool to identify local birds and plants without relying on internet connectivity or paid APIs.

Outcome

The PWA operates entirely offline after the first load, enabling reliable species identification, learning, and logging without incurring API costs or exposing user data.

Workflow

• Load the PWA once while online
• Switch to airplane mode
• Point camera at a bird or plant
• CLIP zero‑shot classifier identifies species or returns "not sure"
• View species card with facts and safety note in English/Tamil
• Ask follow‑up question to local Qwen2.5-0.5B model
• Check or export field log as JSON

Evidence

Madras Field Notes is an installable PWA that turns your phone into a pocket naturalist for Chennai — and it works entirely in airplane mode.

Source

Identify: Point your camera at a bird or plant. A CLIP zero-shot image classifier — running inside your browser, against a curated list of 15 species common to IIT Madras campus and the Pallikaranai marsh — tells you what you're looking at, with an honest "not sure" state below its confidence threshold.

Source

Ask: A Qwen2.5-0.5B open-weight model running locally in the browser tab answers follow-up questions grounded strictly in the species card — it cannot hallucinate edibility or medical advice because it never sees the open internet.

Source

Log: Identifications save to an on-device field journal, exportable as JSON for citizen-science projects.

Source

Service Worker caches the app shell; the Cache API persists model weights; navigator.storage.persist() guards against browser eviction.

Source