You are currently offline

Meta Releases SAM AI Model for Object Identification in Images

Meta, the parent company of Facebook, has introduced a new artificial intelligence model named the "Segment Anything Model" (SAM) along with a vast dataset of image annotations. SAM is designed to recognize individual objects within images and videos, even if it hasn't encountered those objects during its training phase. 

Users can interact with SAM by clicking on objects or providing text prompts, and it will identify and highlight the corresponding items within images. 

This release reflects Meta's ongoing efforts to leverage generative AI technologies for various applications, including content creation and enhancement. Meta's CEO, Mark Zuckerberg, has emphasized the importance of incorporating generative AI into Meta's apps in the coming year.

Meta Releases SAM AI Model for Object Identification in Images

However, as for now the company has made the SAM model and dataset available for download under a non-commercial license. However, those who upload their own images to the accompanying prototype must agree to use the tool only for research purposes.
Post a Comment (0)
Previous Post Next Post