DINOv2
Software / AppMentioned in 1 video
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.
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.
![Best of 2024 in Vision [LS Live @ NeurIPS]](https://i.ytimg.com/vi/76EL7YVAwVo/maxresdefault.jpg)