ordellrb@lemmy.world to linuxmemes@lemmy.world · edit-229 days agoNot Total Recall (1990)lemmy.worldimagemessage-square57fedilinkarrow-up1728arrow-down110
arrow-up1718arrow-down1imageNot Total Recall (1990)lemmy.worldordellrb@lemmy.world to linuxmemes@lemmy.world · edit-229 days agomessage-square57fedilink
minus-squareJackGreenEarth@lemm.eelinkfedilinkEnglisharrow-up12arrow-down5·29 days agoCan you search the screenshots with OCR though? That’s Recall’s main selling point
minus-squareAux@lemmy.worldlinkfedilinkarrow-up58·29 days agoYou can start by running sudo apt install tesseract-ocr and then reading its docs.
minus-squareMacN'Cheezus@lemmy.todaylinkfedilinkEnglisharrow-up4·edit-227 days agoIt appears to be as simple as tesseract <infile> <outfile>. Possibly could even pipe (or tee) the screenshot straight into that and save both an image and a text file in a single command line. So something like this should do the trick: gnome-screenshot -f - | tee /Microsoft/yourPrivacy/$(date +%s).png | tesseract - /Microsoft/yourPrivacy/$(date +%s).txt Skip the database, just use grep to search that directory if you need to find anything. Voilà, homemade Recall.
minus-squareMacN'Cheezus@lemmy.todaylinkfedilinkEnglisharrow-up0·28 days agoYou can use textract to make it even simpler (still uses tesseract under the hood though).
minus-squareR00bot@lemmy.blahaj.zonelinkfedilinkarrow-up18·28 days agoI can’t imagine it’d be that hard to write some code that does that using an existing AI model.
minus-squareMacN'Cheezus@lemmy.todaylinkfedilinkEnglisharrow-up2·28 days agoLlava and Bakllava are two Ollama models than can not only extract text but also describe what’s happening on screen. Using tesseract-ocr, as the other guy suggested, is probably simpler and less resource intensive though.
minus-squareJackbyDev@programming.devlinkfedilinkEnglisharrow-up6·28 days agoThis is a shitpost and not a real suggestion.
Can you search the screenshots with OCR though? That’s Recall’s main selling point
You can start by running
sudo apt install tesseract-ocrand then reading its docs.Fulfills the AI quota 👍
It appears to be as simple as
tesseract <infile> <outfile>. Possibly could even pipe (or tee) the screenshot straight into that and save both an image and a text file in a single command line.So something like this should do the trick:
Skip the database, just use
grepto search that directory if you need to find anything. Voilà, homemade Recall.You can use textract to make it even simpler (still uses tesseract under the hood though).
I can’t imagine it’d be that hard to write some code that does that using an existing AI model.
You’re probably right.
Llava and Bakllava are two Ollama models than can not only extract text but also describe what’s happening on screen.
Using
tesseract-ocr, as the other guy suggested, is probably simpler and less resource intensive though.This is a shitpost and not a real suggestion.