Posted On: Jun 23, 2021

Amazon Textract is a machine learning service that automatically extracts text, handwriting and data from scanned documents that goes beyond simple optical character recognition (OCR) to identify, understand, and extract data from forms and tables. We continuously improve the underlying machine learning models based on customer feedback to provide even better accuracy. Today, we are pleased to announce an accuracy enhancement update to our forms extraction feature. Starting today, Textract more accurately detects the checkboxes and key-value pairs within documents and images. The accuracy is especially improved for invoices and receipts, but customers will see improvements across many documents in finance, healthcare, legal, public sector, and more.

This update will be launched in waves, starting with the Asia Pacific (Sydney), Asia Pacific (Singapore), on June 23rd, followed by Europe (Ireland), Europe (Paris), Europe (Frankfurt), Europe (London), Asia Pacific (Seoul), Asia Pacific (Mumbai) on June 24th, followed by Canada (Central), US West (Northern California), US East (Ohio), GovCloud (US-West), GovCloud (US-East) on June 28th, and US East (Northern Virginia), US West (Oregon) on June 29th.