Microsoft has released version 4.0 of Computer Vision Image Analysis for Azure Service as a preview. In the future, the cloud service will make it possible to call a large number of image analysis functions via a single API. With the 4.0 update, the service gets some new functionality like improved text recognition (OCR, OCR) and human recognition. Computer vision image analysis is of particular interest for professional purposes such as search engine optimization, unobstructed rendering of image content using alt text or automated modification of graphic content.
The biggest innovation in Microsoft’s service is that the Image Analysis API now processes many functions with just one call. This includes automatic OCR, which previously required calling a separate API, image captioning, classification, and tagging. In addition, the API also combines object and, more recently, people recognition.
New features in preview mode
Computer Vision had already perfected optical character recognition before. The new integration into said API should also perform the operation via a centralized user interface and in possible real-time scenarios in the future. In addition, Microsoft Azure has now expanded text recognition to 164 languages, so that OCR now also understands Cyrillic, Arabic, and Hindi. Like this service, the new Spatial Analysis (Spatial Analysis) is initially available as a preview. With the help of these, it should be possible to develop apps that count the number of people in pre-set rooms – for example to ensure social distancing or measure waiting times in exit areas.
Azure Face is also initially available only as a preview and to a limited number of Azure customers only. Using artificial intelligence, this service can “detect, recognize and analyze” human faces. According to Microsoft, this may be necessary above all for identity verification or for obfuscation to make people unknown. Due to the sensitive analysis content, access is initially reserved for those who are among Microsoft’s “managed customers and partners” and can demonstrate their commitment to the company’s “responsible AI” principles.
Interested parties can test the new features of the Computer Vision Image Analysis Preview version via Vision Studio in the browser – without an Azure account, however, only sample images and videos are available from Microsoft. In addition, image analysis is now available via Client-Library-SDK or via REST-API. Microsoft links the documentation at the end of the blog post.
(drowsy)
