3 ms·
Back to the Future of Handwriting Recognition (2016)
- marcind 5y agoI think this challenge remains unanswered to this day in one important way: just about all current handwriting recognition tools assume you're capturing that handwriting digitally (touchscreen, smart pen, smart paper, etc.). That's already impacting the "operational mechanics" mentioned in the article. Vision/light is what the physical and digital worlds have in common. I think this is the only (efficient) way forward.
- dang 5y agoSome past threads: Back to the Future of Handwriting Recognition - https://news.ycombinator.com/item?id=17288454 https://news.ycombinator.com/item?id=17288454 - June 2018 (37 comments) Show HN: Interactive, Visual Explanation of the GRAIL Handwriting Recognizer - https://news.ycombinator.com/item?id=11380711 https://news.ycombinator.com/item?id=11380711 - March 2016 (18 comments)
- Hokusai 5y agoMost of the time my B was recognized as R, as I was not properly closing it at the bottom. It's a very interesting read (and playground) and shows the difficulty of coding hand write text. It goes to the extend to not try that but too create a similar to natural hand writing "code" that is very easy to learn for people that knows hand writing. Nowadays I use a hundred times more "predictive typing" that hand writing. I find it faster, or at least more comfortable. (And being fair, easier to read afterwards)
- gumby 5y ago> It goes to the extend to not try that but too create a similar to natural hand writing "code" that is very easy to learn for people that knows hand writing. The “graffiti” system developed by Palm was an example of this. It was so convenient and easy to use (for those used to Latin scripts) that it was common to see it appear on whiteboards in meetings. https://en.wikipedia.org/wiki/Graffiti_(Palm_OS) https://en.wikipedia.org/wiki/Graffiti_(Palm_OS)