Bulk Photo Import
3-minute read · v1.2 · Collector+
Cataloging a hundred cards, a shelf of bottles, or a wall of comics one-by-one is the kind of task that keeps inventory apps from getting used. Bulk Photo Import takes a single photo of a binder page, a shelf, or a stack and splits it into individual item candidates the AI extracts. You review, confirm, batch-save.
What it works for
- Card binder pages — 9-pocket pages of Pokémon, Magic, sports cards. AI reads the visible card name, set, and number when legible.
- Shelves of bottles — wine racks, whiskey shelves, spirits collections. AI extracts label text + bottle position.
- Comic stacks — fanned out so each title is visible. AI reads issue numbers and titles.
- Sneaker shelves — boxes side-by-side with model name visible.
- Mixed shelves — Funkos, LEGO sets, video game cases. AI labels each detected item with its best-guess category.
How to use it
- Open the + menu, then tap Bulk Photo Import (Collector+ only).
- Take one photo of the shelf, binder page, or stack. Good lighting + items mostly visible (not too overlapped) gets the best results.
- AssetVault sends the selected image and extraction instructions through our self-hosted catalog service to the Google Gemini Developer API. Processing time varies; when it finishes, you get a list of candidates to review.
- Review each candidate:
- Tap to expand and see the AI-extracted name, category, and any metadata.
- Edit anything that's wrong (typo in the card name, mis-categorized bottle, etc.).
- Tap the X to drop a candidate that isn't an item (a price tag, a divider).
- Tap Save all. Each candidate becomes an item in your vault.
Tips for better results
- One row at a time — if you have a 6-shelf wine cellar, photograph each shelf separately. Too many items per photo and the AI can't read the labels.
- Spread items out a bit — items touching or stacked on each other are harder to separate. A binder page works because each card is in its own pocket.
- Even lighting — overhead light, not behind you (your shadow falls on the items). A window during daylight works great.
- Phone parallel to the surface — angled photos cause perspective skew that's harder for AI to read accurately.
- Don't worry about perfection — you're going to review the candidates anyway. The AI does the tedious extraction; you do the final approval.
What it does NOT do
- It doesn't replace AI Photo ID for individual items. If you want maximum accuracy on a single high-value item, use single-photo AI Photo ID — better lighting, better framing, better result.
- It doesn't see through stuff. Items behind other items, or boxes with the labels facing the wall, won't be detected.
- It doesn't grade — bulk import is identification only. For grade estimates, use AI Pre-Grade per-item.
Trial caps
During an eligible 7-day Collector+ trial you get 3 bulk imports — enough to evaluate a few shelves or binder pages. The App Store or Play Store purchase confirmation shows trial eligibility, the billing date, renewal terms, and the cancellation deadline. Paid Collector+ removes the trial counter, but splitting large groups across clear photos still improves accuracy.
Privacy
The selected image and extraction instructions transit our self-hosted catalog service and the Google Gemini Developer API. The catalog service is not designed to retain submitted photo bytes, but Google's terms and retention practices apply after forwarding. AssetVault does not guarantee that every image has EXIF or GPS metadata removed, and any text or sensitive detail visible in the image travels with it. The items you approve are saved in your local vault. Crash diagnostics and purchase verification may still contact their services automatically as described in the Privacy Policy.