Outside Quest: I asked a local Gemma to name birds. It couldn't, so it sends you on a scavenger hunt instead
说明与用途
Outside Quest 使用本地 Gemma 4 模型,通过 Ollama 运行,生成基于地点的寻宝卡,使用严格的 JSON 模式评估照片,并生成步行日志,全部离线完成。
问题
Users want an offline photo scavenger hunt that works without internet or API keys, and that can reliably judge photo content for game scoring.
结果
The app runs entirely offline, generates location‑specific scavenger hunt cards, accurately judges photo content with a strict JSON schema, and produces a walk journal with timeline and route map—all on a laptop with a 6 GB GPU.
工作流
- 1. User types destination and time of walk (≈2 min).
- 2. Gemma 4 writes a six‑item quest card tailored to local climate and prints or displays it.
- 3. User walks, keeping phone in pocket and taking photos only when a quest item is found.
- 4. User returns home and uploads photos to the local server.
- 5. Gemma 4 evaluates each photo against every quest item using a JSON schema of two yes/no questions.
- 6. The app tallies completed quests, displays points, badges, a timeline from photo timestamps, and a small route map from GPS tags.
- 7. All data remains on the laptop; no upload or internet required.
证据
My first idea was the obvious one: a bird and plant identifier that works on the trail with no signal.
A quest only counts if both are true. On a test card (a bird, an insect on a flower, a mushroom, a stormy cloud, leaves up close, something red) and the six test photos, that gave 5 correct matches and no false ones.
No account, no upload, no internet needed. Your photos (and their location data) never leave the computer.
- 记录日期: 2026-10-06T07:55:51.423Z