entry-slick
entry-slick
entry-slick
About 寻隐

An open source, locally running image search app, built on the Core ML model, allows you to use natural language to search for images in the device.

The App currently only has an ios version, if you are an Android device, you can use MEME-Indexer

Hidden Finding is a fully native Core ML model that lets you find photos in your photo album with sentence descriptions.

  • Note: The Mac version currently only supports album search, and does not support search for all images on the disk.

1. Its advantages?

You can use any description you can think of to find a photo, be it a scene (eg "walking in the rain"), a color ("orange lamp in a dark room"), an object's motion ( "Puppy Opens Mouth"), and even the emotion conveyed in the photo ("We're in love").

If you remember taking a certain photo: going camping with your friends, taking a photo with your lover at the beach, but there are too many photos in the album, you can use "a group of people camping in the forest" or "boy and girl Hand in hand by the sea” to find it quickly.

And all of this is done completely offline (without network connection), Xunyin cares about the privacy of your photo album.

2. How do I use it?

When running for the first time, Xunyin needs to build an index for your photo album (remember to authorize it to access all photo albums), which usually takes a few minutes (depending on the number of photos you have), and the processing speed is about 2000 photos/minute.

Building the index is only required once. Once you're done, you can look for photos with any description you like. When you have new photos, you can easily manually update the index to include them in the search.

3. What about the privacy of my photo album?

Hidden is designed to run in a network-free environment, and you can even turn off its network permissions, which means that it is impossible to upload your photos to any place other than the device, so you can trust that there will be no privacy when using it risk.

In the latest version, Xunyin has added the option of "Allow users to set up networking to download iCloud pictures": When photos are only stored on iCloud, you can allow Xunyin to use the official api to download the original image. Of course, this is not mandatory.

4. Are there any equipment and system requirements?

have. Xunyin runs the CoreML model locally, which means that if your device does not support it, the operation will crash or the search results will be abnormal.

The restrictions of the Mac version are as follows:

  • MacOS 13.0 or above system

  • M1 chip or above models

5. Photos stored on iCloud

As mentioned above, Stealth attempts to run without any network connection. If your photos are stored entirely locally, you won't have any problems. When some of your photos are only stored on iCloud (only low-definition cache locally), you can choose to allow Hidden to access the network to download the original photos from iCloud, or you can manually locate the photo in the album based on the date information.

Visit Official Website

https://apps.apple.com/cn/app/id1664361663

Reviews
Show more
白雪公主
I hope the photos can be switched by swiping left and right, and a progress bar will be more intuitive when indexing.
image
Share
伊豆咖啡
The storage space consumption in the latest version has been greatly reduced, and the model loading speed has also become faster.
image
Share
Community Posts
doom guyplan-icon
The more accurate the text description you enter, the more able it is to find the corresponding photo in the album.
image
Share
master chief117
This is a very interesting app. The hit rate is high and it found many unexpected old photos.
image
Share