Vision Transformers

Token Learner

Dynamic module that learns to select and combine the most relevant tokens from the image, reducing computational complexity while preserving essential information.

← Geri