What’s new in Computer Vision OCR

From the creator

On this week's AI Show, Sanjeev Jagtap joins Seth to talk about what's new with Computer Vision OCR https://aka.ms/AIShow/OpticalDoc. Tune in to see Computer Vision updated with industry-leading models built by Microsoft Research. These models are tagging contents in an image with significantly more detail & accuracy, across more languages. Jump to: 00:00 livestream begins 00:15 Seth welcomes Sanjeev 00:48 What is OCR? 02:21 What's new to OCR? Multiple languages in same text line, handwritten and print, confidence thresholds and large documents! 03:55 How Read operations works 05:14 Demo - Read image as text 07:05 Demo - Handwritten classification 09:21 Demo - Mixed languages in same text selected pages extraction 12:10 Seventy three languages supported by Read 13:25 Demo - Text lines in Reading order 15:49 Demo - OCR Applied (Form Recognizer) layouts and tables Learn more: What is Optical character recognition? https://aka.ms/AIShow/OpticalDoc Quickstart: Read client library or REST API https://aka.ms/AIShow/RESTAPIQuickstart How to call the Read API https://aka.ms/AIShow/ReadAPIDoc Install Read OCR Docker containers from Computer Vision https://aka.ms/AIShow/OCRDockerDoc Create a Free account (Azure) https://aka.ms/aishow-seth-azurefree Get Started with Machine Learning https://aka.ms/AIShow/StartML AI for Developers - https://aka.ms/AIShow/AIforDevelopers Azure Machine Learning https://aka.ms/AIShow/AzureML Follow Seth https://twitter.com/sethjuarez Follow Sanje

Published 2021-05-21 · Watch on YouTube ↗
Availability and playback restrictions are controlled by the original channel.