Mark Zuckerberg tests Meta's new AI model on old video of him singing to his daughter Maxima (WATCH)
Mark Zuckerberg shared a throwback video in which he is seen singing and playing a song on a guitar for his daughter, Maxima. While the video surely looks adorable, it has a twist to it. Zuckerberg informed that he tested this video on a new artificial intelligence (AI) model called V-JEPA, and even shared the results for it.
FollowMark Zuckerberg posted an old video announcing the new project named V-JEPA in an attempt to promote Meta's most recent artificial intelligence (AI) model, which learns by watching videos. He posted an old video on Instagram of himself playing a guitar and singing for his daughter Maxima. He stated in the description that he used the AI model V-JEPA to test the movie. This model is described as a "non-generative model that learns by predicting missing or masked parts of a video in an abstract representation space" on its website.
"Throwback to singing one of Max's favourite songs. I recently tested this video with a new AI model that learns about the world by watching videos. Without being trained to do this, our AI model predicted my hand motion as I strummed chords. Swipe to see the results," Zuckerberg wrote in the caption of the post.
Zuckerberg shared two separate videos. In the first clip, he is seen singing and playing the song on a guitar alongside Maxima. He presented the AI model's findings in the second video, demonstrating how V-JEPA filled in the blanks in the film by predicting his hand movements while he played the guitar.
Zuckerberg shared the video just a day back and since then it has accumulated more than 51,000 likes.
According to a blog post by the business, Meta claims to have trained the system using a unique masking technique in which portions of the video were masked in both time and space. This implies that the model has to anticipate not just the current frame but also the following frame in a movie because some frames had blacked-out portions while others had all of their frames deleted. The business claims that the model accomplished both tasks effectively. The model is notable for its ability to anticipate and analyze films up to ten seconds long.
subscribe to Asianet Newsable WhatsApp channel by clicking here.
