Meta, Mark Zucherberg’s own company, recently published an artificial intelligence model. The model can pick out individual objects from within an image.
Meta is Mark Zucherberg’s led-parent company, considered a social giant. The company owns big social media platforms such as Facebook and Instagram.
Moreover, the company has also released a dataset of image annotations. The company claims that it is a fantastic feature of its kind.
The SAM, Segment Anything Model is an advanced AI recognition model that can identify objects within images and videos. In addition, the model can identify them even if it has not encountered them during its training phase.
How Does It Work?
Users can enjoy the features by selecting objects and clicking on them. Or by using text prompts such as the word “dog” or “table” or any other entity.
SAM could point out the desired object in a demonstration by making a square around accurately responding to the written prompt.
Meta said in a blog post, “The model SAM is designed and trained to be promotable so that it can transfer zero-shot to new image distributions and tasks.”
In addition, “we evaluate its capabilities on numerous tasks and find its zero-shot performance impressive. Often competitive with or even superior to prior fully supervised results”.
The company tells the new model SAM could have various applications for content creators. Including the ability to extract images for use in collages or video editing. Moreover, the model will also help in scientific research.
Scientists can take advantage to locate and track animals or objects of interest within video footage of natural occurrences in Space or on Earth.
Moreover, Meta is already using technology for tasks like tagging pictures, removing objectional material, and choosing which articles to recommend to Facebook and Instagram users.
According to the company, introducing SAM would increase access to that technology.
In addition, the dataset and SAM model will be available to access and download under a non-commercial license.
Users who agree that upload their images to an accompanying prototype must also agree to restrict use to research.
Meta has already used other AI tools that not only create brand new content by simply categorising and identifying data like other AI tools.
For instance, the tool that spins up surrealist videos from text prompts mainly generates children’s books illustrations from prose.
According to Mark Zuckerberg, “Incorporating such generative AI “creative aids” into Meta’s app is our priority this year”.
As we know, AI has taken the world to the next level in the last few months. Companies, including high-tech giants like Microsoft and Google, have also incorporated the technology into their search engines.
ChatGPT, AI is another creative innovation of OpenAI, which h sparked a race among tech giants to unveil similar tools.
However, the rapid rise of AI has raised many ethical and societal concerns about technology that can generate fantastic prose or imagery and produces work like humans.
Read more:
Top 10 Metaverse And Web3 Forecasts For 2023
Meta to Cut Off Support for NFTs as Crypto Market Spirals
The post Meta Launches AI That Can Identify Items Within Images appeared first on TechJuice.
from TechJuice https://ift.tt/2CawvkR
0 Comments