Meta unveils V-JEPA AI model that improves training by learning from video

Meta Platform Inc.’s AI research division has released a new artificial intelligence model today that makes crucial steps in AI training that advances learning by interpreting video information similar to the way that humans understand the world. The model, named V-JEPA, or the Video Joint Embedding Predictive Architecture model, works differently than large language models. […]

The post Meta unveils V-JEPA AI model that improves training by learning from video appeared first on SiliconANGLE.

advances ai ai model ai research ai training architecture artificial artificial intelligence embedding humans information intelligence jepa joint embedding predictive architecture meta meta platforms inc. platform predictive research the-latest training video v-jepa world yann lecun