09-24, 11:55–12:30 (Europe/Lisbon), Auditorium
Medical Charts contain information which is documented in traditional Handwritten format which cannot be processed by NLP. Let's dig into how to build an API which detect handwritten information using FastAPI.
Talk Description:
Introduction
We will discuss on the importance of detecting handwritten text in medical charts with some examples.
Basics
We will mainly discuss about Object detection frameworks and go brief on Faster RCNN framework, how we implemented it using Detectron. We will then deploy the model through FastAPI.
Methodology
We will go over through whole Data Cycle of Model. It includes curating, annotating and training the dataset and modelling it and then packaging it into an API.
In the end we will talk about the results and takeaway from the project.
Prashant is a data scientist, skilled in application of machine learning techniques for solving data based problems. Currently, a Data Scientist working on Computer Vision Applications.
He is an active community guy and loves to share and learn as much he can by organizing Meetup and networking with individuals.