GLIP: Introducing Language-Image Pre-Training to Object Detection | Towards Data Science

Grounded Language-Image Pre-training by L. H. Li et. al.