Abstract: A vision transformer (ViT) is a deep learning model that performs one or more vision processing tasks. ViTs may be modified to include a global task that clusters images with the same concept together to produce semantically consistent relational representations, as well as a local task that guides the ViT to discover object-centric semantic correspondence across images. A database of concepts and associated features may be created and used to train the global and local tasks, which may then enable the ViT to perform visual relational reasoning faster, without supervision, and outside of a synthetic domain.

Abstract: A spraying assembly, including an atomizer, a sprayer and a rubber plug, and an atomizer inner cavity and an atomizer channel located below the atomizer inner cavity are provided in the atomizer, an atomizer end (130) is included in a lower part of the atomizer, and a side wall of the atomizer end is provided with at least one opening; the rubber plug is located at a bottom end of the atomizer inner cavity, an upper end of the rubber plug is provided with an umbrella-shaped boss, an annular cavity is formed between a lower end of the rubber plug and an inner wall of the atomizer, the rubber plug is provided with at least one longitudinal notch along a circumferential direction, and the bottom end of the atomizer inner cavity is provided with at least one notch; the sprayer includes a nozzle.