Oddbean new post about login | logout Are there any good open source handwriting OCR softwares out there? #asknostr #ocr #foss
Tesseract works. Not sure if its "good" compared to SOTA. There's a port of it to Android on F-Droid called simply "OCR".
Ooof. That's not great. My hand writing is not great, but... Oof.
Hah! My cursive is indecipherable even to me the next day, but I can print just well enough that the bots can get it 90% right. Still quicker for me to type, even on phones...
No idea what ocr means, but I did search because you asked twice. https://mindee.github.io/doctr/ https://github.com/Calamari-OCR/calamari https://github.com/mittagessen/kraken
But it doesn't seem to work for handwriting, at least in that one implementation. I should look at what's available for Linux instead of android, which is another annoying thing I'll grumble about to myself...
I get the hype aversion. I get the 'technology' aversion - but if you want to recognize handwriting in an image, you're not going to do it by explicitly building rules from the ground up, it just doesn't work that way. "All" machine Learning" and AI (deep learning) is, is a defined process to minimize errors, and it uses numbers to do so. The specific numbers are essentially the rules that have been 'found' by the algorithm that minimizes error, which you don't necessicarily care about because you just want the right answer. You could simpify all of AI and ML algorithms to playing advanced games of "hot and cold". - You have a bunch of pixel values that represent text, - You make a guess "this looks a lot like an L, an i, or 1, but it looks most likely to be a 7, so I'll say that this image is a 7" - turns out the correct answer was "T" but they write like a chicken, so I'll have to make an adjustment for next time and hopefully be correct then.
Few understand this. AI is like gunpowder: master it and use it to proliferate your values or get subjugated by those who do.
I am just not equipped to handle AI as a tool. I still hurt myself and others with it by fumblefarting around with it. So, I will do without or just not do at all.
Understood it can be intimidating. This is an area I’m passionate about I’m building @CASCDR and I gave a really straightforward “entry point” to AI talk here: https://m.youtube.com/watch?v=BJv2McQnX6I . First ten mins are about demystifying AI which is not as complicated as nerd charlatans need you to believe it is.
And I'm mindful that you have space constraints, otherwise the cost of a six-year-old gaming rig would be reasonable. Last night I was looking at a $100 RISC-V SBC and plotting how I could get a real graphics card onto it for my container-cabin. There are mini-PCI adapters, but powering it would require a beefy external supply. And I don't think CUDA for RISC-V is a thing yet :(
I am just not equipped to handle AI as a tool. I still hurt myself and others with it by fumblefarting around with it. So, I will do without or just not do at all.
Understood it can be intimidating. This is an area I’m passionate about I’m building @CASCDR and I gave a really straightforward “entry point” to AI talk here: https://m.youtube.com/watch?v=BJv2McQnX6I . First ten mins are about demystifying AI which is not as complicated as nerd charlatans need you to believe it is.
Think you're not giving yourself enough credit, while simultaneously giving the data scientists too much credit in how well they understand the tools they build 😅 https://imgs.xkcd.com/comics/machine_learning_2x.png
