Tech News

Meta launches AI model that identifies objects in images and videos

Meta announced on Wednesday the Segment Anything Model (SAM), its new artificial intelligence (AI) model that can identify individual objects in images and videos. 

According to the company’s AI research blog post: 

  • In short, the SAM identifies which image pixels belong to a certain object, it is like a powerful refined filter; 
  • For example, in a demo video, typing the word “cat” caused the tool to draw boxes around each of the multiple cats present in an image; 
  • In another example, while filming, the AI ​​identified each object that appeared on the recording screen; 
  • The tool can also separate the same object by layers or highlight them in 3D format.

We intend to democratize segmentation by introducing the Segment Anything project: a new task, dataset and model for image segmentation, as explained in our research paper. We are releasing our Segment Anything (SAM) general model and our Segment Anything 1-Billion (SA-1B) mask dataset, the largest segmentation dataset ever, to enable a broad set of applications and promote additional research on underlying models for computer vision. 

Explained the AI ​​team at Meta, in an official statement .

According to big tech, the features are based on the type of generative AI popularized by ChatGPT, which creates entirely new content rather than simply identifying or categorizing data. For the company’s Chief Executive, Mark Zuckerberg, incorporating these “creative aids” into the company’s applications is a priority for this year. 

Meta launches AI model that identifies objects in images and videos
Object selection ‘dog’ in demo. Image: disclosure/Goal

It’s worth noting that Meta already uses SAM-like technology internally for things like photo tagging, moderating banned content, and determining which posts to recommend to Facebook and Instagram users. The launch of SAM to the public and researchers aims to expand access to this type of technology. 

The SAM model and dataset are available for download under a non-commercial license. 

AI advancing pause 

Meta’s announcement of its new AI model for images comes as the tech market is grappling with a conflict over whether to slow the race for artificial intelligence (AI) products as powerful as GPT-4 (the latest version of OpenAI’s ChatGPT). 

In an open letter  issued by the Future of Life Institute, leading names in the  tech sector , such as Elon Musk, reached a consensus calling for a pause in the rise of AI systems, stating that accelerated advancement, without safety regulation, is a potential risk for the humanity. In the case of ChatGPT, text generative AI, the main concerns are disinformation and cybercrime.

Related Articles

Back to top button