With the introduction Intelligence Artificial in the daily life of each of us, even the former Facebook has decided to take a few steps. It couldn't help but do it, actually, being one of the largest technology companies in the world. Indeed Meta presented SAM. It is artificial intelligence model capable of locate individual objects within an image, along with an image annotation dataset that the company says is the largest ever of its kind.
Meta introduces SAM a new image segmentation task, model, and dataset trained to be timely
The company's research division said in a post on the blog that his Segment Anything Model, or SAM, could identify objects in pictures and videos even in cases where it hadn't encountered those items during its training. Using Meta's SAM, objects can be selected by clicking on them or writing text messages. In one demonstration, writing the word “cat” prompted the tool to draw boxes around each of the different cats in a photo.
Meta's publication of the largest image annotation dataset ever created provides a advantage competitive significant in the development of artificial intelligence, making information easily accessible for training AI models. In this way, AI models can learn not only from human input, but also from any existing photo or image.
An example of how Meta's SAM model can be used is real-time image and video selection for the dealer online, allowing you to select products or manufacturers that match the photo or video you are viewing. This opens up a number of opportunities, for example the promotion of products in specific contexts for their use. However, the side effects of using AI, such as eliminating human employment, the decline in data security and lack of democratic oversight, will need to be carefully monitored and mitigated.
This is something we could see with the matter of block in Italy of ChatGPT following the recovery by the Privacy Guarantor of the system designed by OpenAI