Azure Cognitive Service for Vision has released for public preview the next generation of computer vision capabilities to help Developers create cutting-edge, market-ready, responsible computer vision applications across industries. Customers can now seamlessly digitize, analyze, and connect their data to natural language interactions and unlock powerful insights from their image and video content to support accessibility, drive acquisition through SEO, protect users from harmful content, enhance security, and improve incident response times. It enables significant improvements to image captioning and groundbreaking customization capabilities with few-shot learning. Until today, model customization has required large datasets with hundreds of images per label to achieve production quality for vision tasks. Now Cognitive Services for Vision is capable of recognizing millions of object categories out-of-the-box, which makes features like captions rich with details and sematic understanding. Custom models can achieve high quality when trained with just a few images, lowering the bar for creating computer vison models that support challenging use cases. 

Learn more about Azure Cognitive Service for Vision

Try it out in Vision Studio: Vision Studio (azure.com)

Leave a Reply