Apple has recently released Ferret, a Multimodal Large Language Model (MLLM) AI model on GitHub. This groundbreaking model, which was actually released on October 30th but just recently discovered and reported, distinguishes itself with its ability to extract and analyze visual data in any given shape or area. For example, it can identify the tail of a cat from an image or separate two overlapping animals in a picture.
What’s particularly interesting about Ferret is that it has been published as non-commercial open-source software, allowing researchers to use it for further development and experimentation without seeking any financial gain. This move is seen as a positive step by AI researchers, as it is quite rare for Apple to release research projects of this nature that have the potential to be developed into marketable products.
In summary, Ferret is a revolutionary MLLM that excels in its capability to refer and ground visual information with remarkable precision. Its recent release in an open-source format marks a significant milestone in Apple’s approach to research and development.
TLDR: Apple has unveiled Ferret, an AI language model with unrivaled prowess in analyzing visual data. Notably, Ferret can accurately identify specific regions within an image and has been made available in a non-commercial open-source format, allowing researchers to explore its potential without any profit-seeking motives. This move signifies a notable stepping stone in Apple’s research endeavors.
Leave a Comment