Hi HN,

I built a native desktop app that helps you quickly search your local images using either a text description or an image, completely offline, and no GPU required. It would be great to see what you guys think about it and have a bit of feedback.

Features:

- Text-based search: You can search your images with textual descriptions like "beer", "pizza", "a man walking on the beach", etc.

- Image-based search: You can drag and drop an image and the app will show you visually similar ones. You can also use the result images as input, which creates an interesting way of navigating your images.

- Multilingual: works out of the box in English, Spanish, French, Japanese, and other languages. There's no need to change any setting, for example you can write something in English and then write something in Japanese, and both searches will work seamlessly.

- Private and offline: nothing leaves your machine, as all the processing is done locally. At the beginning, embeddings of your images are generated and stored in an sqlite database, and then the searches are run directly on that local file.

- Simple UI: One of the main things I wanted to do is to make this super easy to use. You just drag and drop folders to index them, and an image to search with it. You can also just type at any time and that's your text search. Also provided are more standard ways of interacting with the software like keyboard navigation, file pickers, etc.

How to get it:

I published a free demo version for Windows and macOS that lets you try all these features with up to 200 indexed images. No limits on how many times you search or how long you use it. You can keep this for free, forever. If you're interested in it, you can get it at: https://samontab.itch.io/find-that-image

I made this to scratch my own itch, I have tons of local images and wanted to search them, but I usually found that the best solutions to this were cloud-based. Not much was available for offline usage, so I thought I could make something useful for me, and maybe others would find it useful as well. I'd love to hear your feedback.

Thanks for reading!