Skip to content
AIMarketCap

V-JEPA

Model
MetaFamily V-JEPAlegacy

Video Joint Embedding Predictive Architecture model that learns visual representations by predicting in latent space.

Released
Feb 15, 2024
Context window
—
Modality
video
Open source
—
API available
—
Releases
1

Releases

1
Feb 15, 2024V-JEPA releasedV-JEPA · 1

Meta released V-JEPA, a video model trained by predicting missing representations rather than pixels.

Meta AI Open weightsMetaV-JEPA