Распознавание изображений для бизнеса

2020. Microsoft Azure AI now describes images as well as humans


A new neural network model for generating image captions is now available to consumers through Computer Vision in Azure Cognitive Services, which is part of Azure AI, allowing developers to use this capability to improve the accessibility of their services. According to the study, when evaluated on nocaps tests, the AI ​​system generated more meaningful and accurate captions than humans did for the same images. The new model is already used in the Seeing AI app and will be available later this year in Microsoft Word and Outlook for Windows and Mac, as well as PowerPoint for Windows, Mac, and the web. Automatic captioning helps users access the important content of any image, whether it's a photo returned in search results or an illustration for a presentation.