Meta Launches its AI Model ‘SAM’ for Object Recognition in Images
Meta Launches its AI Model ‘SAM’ for Object Recognition in Images

Meta Launches its AI Model ‘SAM’ for Object Recognition in Images

Meta Launches its AI Model ‘SAM’ for Object Recognition in Images

Written by?Sanjay Kumar

Meta has unveiled a new artificial intelligence model that can detect particular objects present in an image. Alongside the model, Meta has also published a dataset of image annotations, which they claim is the most extensive of its kind to date.

Meta’s research division announced in a recent blog post that they have created a sophisticated object recognition model named the Segment Anything Model (SAM). The model has the ability to detect items in images and videos, including those that were not present during its training process. SAM permits users to highlight objects by either clicking on them or providing text prompts like “cat.” During a demonstration, SAM accurately drew boxes around several cats in a picture in response to a written prompt.

Meta

Technology Company

  • Headquarters: Menlo Park, California, United States
  • Founded: February 2004, Cambridge, Massachusetts, United States
  • CEO: Mark Zuckerberg (Jul 2004–)
  • Revenue: 8,596.5 crores USD (2020)
  • CTO: Andrew Bosworth
  • Subsidiaries: WhatsApp, Novi Financial, Inc., Giphy, MORE
  • Founders: Mark Zuckerberg, Eduardo Saverin, Andrew McCollum, Dustin Moskovitz, Chris Hughes

SAM-like technology has been utilized by?Meta ?for internal activities like labeling photos, screening inappropriate content, and proposing posts to Facebook and Instagram users. Meta has expressed that the availability of SAM will broaden the reach of this advanced technology beyond its internal processes.

Under a non-commercial license, the SAM model and dataset have been made available for download by the company. Nevertheless, individuals who choose to upload their own pictures to the accompanying prototype must consent to utilize the tool solely for research objectives.

“In the future, SAM could be used to help power applications in numerous domains that require finding and segmenting any object in any image. For the AI research community and others, SAM could become a component in larger AI systems for a more general multimodal understanding of the world, for example, understanding both the visual and text content of a webpage. In the AR/VR domain, SAM could enable selecting an object based on a user’s gaze and then “lifting” it into 3D,” said Meta in its blog post.

Quick Read:-

要查看或添加评论,请登录

社区洞察

其他会员也浏览了