Sidequest: A 100% Offline Vision‑Language Scavenger Hunt Where the Screen Is the Shortest Part of the Walk

Описание и применение

Sidequest – это полностью автономное приложение для охоты на предметы на природе, использующее открытый модель CLIP для мгновенной проверки найденных объектов, сохраняя экран как кратчайшую часть прогулки.

Задача

Parents give children digital games that replace outdoor exploration with addictive screens, while paper scavenger hunts cannot verify discoveries.

Результат

Sidequest delivers a fully offline, on‑device vision‑language verification workflow that keeps screens minimal, protects privacy, and provides instant feedback to children and parents during outdoor scavenger hunts.

Рабочий процесс

• Present a daylight‑readable naturalist clue on the phone
• Child walks and observes the environment with the screen dark
• Child taps camera to capture a photo of a candidate object
• On‑device quantized vision model evaluates the image in 43 ms and displays a Blaze progress bar
• If Blaze reaches 100 % a field note appears and the child can continue
• Parents may tap "Count it!" to override ambiguous matches
• After all 5 stops a printable field journal is generated locally

Доказательства

Sidequest does not use artificial intelligence to glue eyes to glass. It uses an open-weight vision-language model running inside the browser as a quiet, objective trail referee:

Источник

In just 43 milliseconds, an on-device quantized vision model evaluates whether the image genuinely matches the clue against competing distractors.

Источник

The entire codebase is open-source under the Apache-2.0 License: https://github.com/Shreyansh00987/Sidequest

Источник