Artificial IntelligenceNews

ABBYY’s AI-powered Real-Time recognition SDK helps developers add ‘instant’ text capture functionality to mobile apps

Voice typing is considered as Alpha, till date; where one doesn’t have to type. So, here’s for those who do not even want to waste their energy in speaking out. Abbyy, an organization for document recognition, data capture, and language processing, has released an AI-powered Real-Time Recognition (RTR) SDK.

You don’t have to speak; it speaks for you, that too in 63 different languages!

With the RTR SDK, Abbyy’s mission is to take data capture to an altogether different level. It allows developers to capture content, extract information from it and make it actionable using Artificial intelligence technologies. Developers can now make use of such an SDK to incorporate real-world document data into their apps instantly. The SDK is available to all developers and is ready for integration with existing or new applications for Android or iOS apps. In addition to this, a developer toolkit is also available, which supports an easy integration with the code samples and the quick-start guides.

By using a smartphone camera, the solution is able to capture texts in real-time, be it within a document or over an object. Jupp Stoepetie, Managing Director at ABBYY says, “With real-time recognition, app users can effortlessly enter data from printed sources, documents, and bank cards”. With the RTR SDK, any application is capable of taking data entries virtually from a printed source. This allows fast, easy, and an accurate way of text recognition and classification.

Presently, the Real-Time Recognition SDK is integrated within applications designed for different sectors such as e-commerce, finance, logistics, and government. Having such a solution within applications for instant data capture significantly speeds up various time-consuming tasks such as

  • a registration,
  • opening an account,
  • entering credit/ debit card details or a promo code,
  • applying for a loan, and so on.

The most important feature to note is, the SDK allows you to do all the above without any security implications. As the text captured doesn’t require to be photographed, or saved in memory or onto any cloud or external servers, it best suits organizations which have improved data security standards.

Jupp Stoepetie states,“The technology is very well-suited for processes that require compliance with security and privacy rules, as no images are sent to the server or stored on the device”.

Below is an overview of the key features that ABBYY’s real-time recognition SDK offers:

  • It includes an incredible OCR (Object Character Recognition) technology, backed with intelligent algorithms in order to recognize texts even from within a live video with a high accuracy rate.
  • Real-time extraction of texts from natural scene objects, documents, invoices, and even from the screens (desktop, tablets, and so on) is now possible.
  • One can now extract texts directly using the smartphone’s preview screen rather than taking a picture of the text and re-typing it manually.

When asked about the future of the RTR SDK, and what the developers can expect, Stoepetie said that ABBYY is working with neural networks to improvise the accuracy of the text recognition within the applications via the SDK. He further added that ABBYY’s would be looking into the opportunities provided by the new AI chips within mobile devices. Such AI-powered devices can go hand-in-hand with the intelligent text capture by the RTR SDK enabling a continuous improvement based on user’s feedback. Two of the popular mobile processors that allow on-device AI and machine learning are Apple’s iPhone neural engine and Huawei Mate 10 neural processing unit.

For a more detailed overview of the real-time recognition SDK, you can visit the official link here.


Savia Lobo

A Data science fanatic. Loves to be updated with the tech happenings around the globe. Loves singing and composing songs. Believes in putting the art in smart.

Related Articles

Leave a Reply

Your email address will not be published. Required fields are marked *