Robot man on a park bench with a laptop
Robot man on a park bench with a laptop
Robot man on a park bench with a laptop

Meta Debuts AI Tool for Identifying & Separating Objects in Images

BKMT Staff

Apr 6, 2023


After the Metaverse debacle, Meta shifted focus to AI as its next stage of development. Among its latest projects is the "Segment Anything" model, unveiled by the social media giant on Wednesday.

Segment Anything is a tool that enables users to identify particular objects in an image with just a few clicks. Though currently in demo mode, Meta claims that Segment Anything is already capable of analyzing every pixel in a photo and distinguishing each item in the picture, allowing one or more objects to be separated from the rest of the image.

"Segmentation - the process of identifying which pixels in an image belong to an object - is a fundamental task in computer vision and finds application in a wide range of domains, from analyzing scientific imagery to editing photos," Meta stated in a post announcing the new model.

Meta explained that creating an accurate segmentation model for specific tasks requires specialized technical expertise, along with access to AI training infrastructure and large volumes of carefully annotated in-domain data.

Artificial intelligence systems that create text, images, or other media in response to prompts are known as generative AI. Examples of this technology include OpenAI's ChatGPT and Midjourney, a digital art platform.

Meta has revealed that the Segment Anything AI system was trained using a dataset of over 11 million images. The company has now made Segment Anything available to the research community under the permissive open license, Apache 2.0, which can be accessed via the Segment Anything Github.

After laying off over 10,000 employees following the termination of its Instagram NFT project, Meta announced in February that it was shifting its focus away from the metaverse and towards other products, including artificial intelligence. A new product group dedicated to generative AI was created in the wake of this shift.

However, global leaders have grown increasingly apprehensive about the advancement of artificial intelligence, with concerns about user privacy and safety leading to open investigations into the technology. Italy has even gone so far as to ban the popular chatbot, OpenAI's ChatGPT. While AI continues to advance, it remains to be seen if tech giants can build trust within the online community given its track record on privacy.