Skip to content

Details

In this week’s session we will be covering DINOv1 (Emerging Properties in Self-Supervised Vision), a self-supervised training method developed by Meta.

During the talk we will discuss the model and training architecture, walk through the code, and explore results on downstream tasks

Paper: https://arxiv.org/abs/2104.14294

Related topics

Events in Boston, MA
Machine Learning
Computer Science

You may also like