Hacker News Reader

Comment by johnisgood | Hacker News Reader

I have a privacy concern. Does this send the image (or the URL to that image) to their server for processing, or does it happen in my browser without their server receiving any information regarding the sites I visit, or the images I load?

View on HN

"By default, when you begin selecting text, it sends a secure HTTPS request containing the URL of the specific image and literally nothing else (no user tokens, no website information, no cookies or analytics) and the requests are not logged. The server responds with a list of existing translations and OCR languages that have been done. This allows you to recognize text from an image with much more accuracy than otherwise possible. However, this can be disabled simply by checking the "Disable Lookup" item under the Options menu."

Is the source code of their server available anywhere? Assuming I have it, it would be fairly easy to modify the extension to use my server instead of theirs, right?

(creator here)

If you enable the "Disable Lookup" setting from the context menu the extension doesn't communicate with any servers at all— everything is processed with a Javascript/WebAssembly OCR engine bundled with the app— it just may be of a sightly lower quality.

Hey! Is there some reason why its not working on some photos where the cursor is replaced with a magnifier tool?

This does not answer my question though. Is the source code for the server available? If it is more accurate, and is open sourced, then I would like to run my own and make the extension connect to my server.

I wonder why couldn't this be done using all client-side javascript implementations? I wouldn't mind waiting a few seconds to select versus sending that data to a remote server (and paying for that server through ads or subscription).

(creator here)

That's exactly what pressing "Disable Lookup" does.

We started a fairly popular Emscripten/WebAssembly port of the Tesseract OCR engine specifically to advance the state of client-side javascript OCR engines: https://tesseract.projectnaptha.com

However, there's a limit to the kind of performance that you can get out of client side javascript OCR engines— with a shared server you can get better recognition on certain popular images.

They are the same people behind the javascript port of the tesseract OCR engine: https://github.com/naptha/tesseract.js

Here's an online notebook for trying it: https://observablehq.com/@tmcw/tesseract-js-v2-alpha

I haven't used this extension in a while but IIRC there was an option to do all the OCRing locally.

Reply on news.ycombinator.com