Team Weill Cornell Medicine
team icon
Kelly Wang CM
team icon
Aaron Yingxiang Lu CS
team icon
Evan Yates MBA
team icon
Young Sang Choi HT
team icon
Rohun Triphathi CS
outlink icon Sprint 1
outlink icon Sprint 2
outlink icon Sprint 3
Weill Cornell Medicine asked
How might we recognize handwritten medical charts and convert them to searchable text?
Handwritten text recognition is an unsolved problem at the moment. Every day, there are millions of handwritten documents produced, either from daily logs written by nurses, doctors or notes taken by lawyers, etc., and all of these information is hard to be gathered and sorted due to its nature. So what we propose is a website that could provide all kinds of handwritten text transcription services, leveraging computer vision and machine learning techniques. On that website, the users could simply choose the domain of their scanned documents (due to differences in dictionaries and fonts) and upload it. By feeding the images of scanned documents into neural networks, we will be able to localize and recognize the text on it, and return them into searchable text.