Meta releases AI model Segment Anything: capable of recognizing objects in images

Meta发布AI模型Segment Anything:可识别图片中的对象

Meta has released a new AI model called the Segment Anything Model (SAM). SAM can be used to identify individual objects in images and is suitable for applications that require object detection and segmentation in images. It can also be used to enhance AR and VR applications, such as selecting objects based on user gaze. Meta has adopted a dataset called Segment Anything 1-Billion mask (SA-1B), which it claims is the largest segmentation dataset ever created. The dataset will be available for research purposes under a permissive open license (Apache 2.0).

Meta has already applied this technology on platforms like Facebook and Instagram, helping users tag photos or set up content restrictions, as well as using algorithms to provide personalized post recommendations. Meta has also launched a model demonstration website where users can experience the technology without registering: https://segment-anything.com/

Meta releases AI model Segment Anything: capable of recognizing objects in images
(0)
Harley Norris的头像Harley Norris

发表回复

您的邮箱地址不会被公开。 必填项已用 * 标注