Transferring only the attention maps of a pre-trained ViT recovers the accuracy gain of full fine-tuning on ImageNet-1K classification.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2024 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
On the Surprising Effectiveness of Attention Transfer for Vision Transformers
Transferring only the attention maps of a pre-trained ViT recovers the accuracy gain of full fine-tuning on ImageNet-1K classification.