Photo Finder App - Queryable 4+
Album Search & Cleanup Storage
Yeonni Lee
-
- £4.99
Screenshots
Description
Queryable is an MacOS app that allows you to find a photo in albums use a text description, completely offline.
- Note: Currently, only photos in photo albums are available for search.
1. What's the highlights ?
- You can use any description you remember to find the photo. For example, a scenario described in your mind (e.g. "photos of our gathering"), a tone (e.g. "kite in blue sky"), an object (e.g. "a dog, a girl, flower"), or even an emotion depicted in that photo (e.g. "We fell in love").
- Queryable runs completely offline. This allows you not to worry about anyone invading your album privacy.
2. How to use Queryable ?
* Build index
The first time you run Queryable, it will build index for all your photos (remember to grant the app access to all your photos), which usually takes a few minutes (depending on the number of photos you have) and processes at a rate of about 2000 photos per minute.
This step is performed only once.
* Search as you like
Once that's done, you can start searching for anything you want.
* Manually update index
When you have new photos, manually updating the index will make them searchable.
3. Are there any privacy or security issues?
Rest assured, Queryable is designed as an offline app and will never request network access, thereby avoiding any privacy issues. (You can turn off its network permissions if you wish.)
- update: for version>=1.1.2, Queryable now gives users the option to choose whether to use wireless data to fetch photos from iCloud. This feature is not mandatory.
As a Chinese indie developer who values privacy, it is very difficult to promote product in English speaking regions. But luckily, when it comes to this app, you don't have to believe me, you can trust science.
4. Any requirements for the iPhone model?
Queryable runs CoreML locally, which means if your iPhone does not support CoreML, you can't use it.
The limitations are:
* MacOS 13.0 or above
* M1 chips or later models
5. Issues with iClould
Since it cannot connect to a network, Queryable can only use the cache of the low-definition versions of your local Photos albums. This does not affect search accuracy, except that you cannot view the original image in the search results.
It's a compromise between privacy and security and convenience. An app that has access to your Photos and the network is a disaster. I hope you can understand.
6. What if it crashes when building an index ?
Make sure you're MacOS 13.0, M1 chip or above. Turn off low battery mode and kill the memory-hungry app, then try building the index again.
If there are still problems, please contact: myfancoo@gmail.com
Preview page created with https://screenshots.pro/.
What’s New
Version 1.2.5
- adjust the UI to fit MacOS
- add search by photo feature
Ratings and Reviews
Works well, a little basic right now
The searching is great, but the interface could do with several improvements:
Sort results (by date would be good, choosing newest/oldest)
Show/filter by location
Show info about photo by default (there’s space onscreen to show the image date, why do I have to press i to see it?)
Show photos taken around the same time as the one selected.
Better photo app integration (must be some way to view the image direct in the photo app?)
Developer Response ,
Thanks for making so many suggestions! Regarding sorting by date and geographic location, this is already in todo, but Queryable can't get the user's specific city (converting lat/long to city requires a network query).
The current design is mimicking Apple albums, but it could indeed become more intuitive, and there are many UI aspects of Queryable that need to be optimized.
In addition, Apple does not allow third-party apps to jump to the album, I will try to make it more convenient in subsequent versions.
Academically interesting
Nice idea from an academic perspective but fails to meet the bar for a paid app and isn’t really better than the Photos app.
* the UI is rudimentary at best
* the model takes a long time to long load so the benefit of fast search is lost due to this wait time. Consider that iOS might boot your app out of memory any time, especially when switching between a lot of apps. When you next foreground the app the model has to reload. Developer: there’s a simulator setting that forces iOS to be aggressive about this for testing purposes
* it’s not even that fast, 10,000 photos on iPhone 14 Pro Max, can take a couple of seconds to execute
* can’t identify people by name (it will find people but the wrong people - how could it know anyway?)
* can’t identify locations accurately (it might find somewhere you search for by name but entering the name of places I’ve never been still yields results)
Edit: thanks for the response. Tweaked my review slightly.
Developer Response ,
Thanks for the review.
1.Yes, the UI is rather rudimentary. I thought about making it prettier, but that's not something I'm good at and I'll try to do better afterwards.
2.Loading the model only takes a long time the first time, after that it will be fast. Maybe this is unavoidable for large models.
3.The speed is very unusual. On my iPhone 12 mini, a single search for 10,000 photos takes less than 1 second. Maybe it's faster when you turn off low battery mode, or close apps with high CPU usage in the background.
3."identify people by name" means that it needs to know your contact information first, and I think it is very insecure for a third party to get this information.
4.The lag/long information of the photo could be get easily, but converting it to a specific city requires a network query. So it is not possible to get the city name offline, and to search by city. I'm trying to figure out how to parse out the city name in the offline case.
I hope this can answer your question, thanks again for your comment!
Pretty incredible app
Couple of things that would be nice improvements.
Incremental index building on the first launch. It took a while to index all my photos because the app kept getting killed, and then needed to start the indexing process from scratch every time.
Automatic background index updating while the app is open. Feels a bit silly having to press a button to index new photos, would be nice if the app just did that without promoting me first.
Developer Response ,
Thank you for your comment! My considerations are as follows.
1. incremental build indexing is already done in the app, but not for every single one (because doing so I/O consumption is high), I will try to reduce the incremental batch to preserve the current progress.
2. Building the index requires loading another model (img encoder), the load itself usually takes 3-5 seconds, but building the index for 100 photos afterwards only takes 1-2 seconds, so if the index is automatically built once for each new photo, the app will open far slower than it does now.
So it's actually more advisable to build it after you've accumulated dozens or hundreds of photos, but you're right, that button do look silly now.
App Privacy
The developer, Yeonni Lee, indicated that the app’s privacy practices may include handling of data as described below. For more information, see the developer’s privacy policy.
Data Not Collected
The developer does not collect any data from this app.
Privacy practices may vary based on, for example, the features you use or your age. Learn More
Information
- Provider
- Yeonni Lee
- Size
- 280.9 MB
- Category
- Photo & Video
- Compatibility
-
- iPhone
- Requires iOS 16.0 or later.
- iPad
- Requires iPadOS 16.0 or later.
- Mac
- Requires macOS 13.0 or later.
- Apple Vision
- Requires visionOS 1.0 or later.
- Languages
-
English
- Age Rating
- 4+
- Copyright
- © Ke Fang
- Price
- £4.99
Supports
-
Family Sharing
Up to six family members can use this app with Family Sharing enabled.