Vision Transformers (ViT)

Self-Supervised ViT Pre-training

Unsupervised training methods like DINO or MAE leveraging the Transformer structure to learn without annotations.

← Wstecz