DINOv2

Software / App

A self-supervised foundation model trained purely on image data, which learns fine-grained visual features and is used to identify images difficult for LLMs.

Mentioned in 2 videos

Videos Mentioning DINOv2

Best of 2024 in Vision [LS Live @ NeurIPS]

Latent Space

A self-supervised foundation model trained purely on image data, which learns fine-grained visual features and is used to identify images difficult for LLMs.

Stanford CS25: Transformers United V6 I From Representation Learning to World Modeling

Stanford Online

A pre-trained encoder used in the DINO model, which demonstrated that a frozen DINOv2 encoder could provide meaningful abstractions for planning.