Touch Grass — A Bird ID App That Works When There's No Signal
Описание и применение
Touch Grass – это прогрессивное веб‑приложение, которое использует открытые модели AI для идентификации птиц, растений и насекомых полностью на устройстве, без передачи данных на сервер.
Задача
Identifying bird calls typically requires uploading audio to a cloud API, exposing location and usage data.
Результат
The app enables quick, private bird identification offline, preserving user privacy and allowing use in low‑signal environments.
Рабочий процесс
- Record a 3‑second audio clip
- Convert clip to a mel spectrogram locally
- Run BirdNET‑ONNX model in a Web Worker
- Return species prediction without any network call
Доказательства
Touch Grass is a Progressive Web App that runs open‑weight AI models entirely inside the browser.
Record 3-second audio clips, identify species via BirdNET‑ONNX (~20MB, quantized, 3000+ species), all on‑device.
After initial model downloads, no internet required. Service worker caches the app shell; IndexedDB stores observations and model cache.
- Дата записи: 2026-10-07T07:55:54.863Z