Meta's AI Tools Revolutionize Image Segmentation and Recognition.
Meta has released its new AI-based Segment Anything Model, capable of identifying and separating specific objects in images and video, to anybody by making its new software open source. The system is described as “promptable” meaning it can receive input such as users’ gaze in a VR headset or through clicks and even text. The company is still working on making a public release for its ChatGPT competitor called LLaMA, even though it had already leaked online.