inference efficiency 1relevance anchoring 1training-free methods 1vision-language models 1visual token pruning 1
From the 1 of 2 linked papers with an AI index.
2 papers
cs.CV2026
AnchorPrune: Relevance-Anchored Contextual Expansion for Visual Token Pruning
Kyuan Oh, Bumsoo Kim
The paper proposes AnchorPrune, a training‑free method that first selects a relevance‑anchored set of visual tokens and then expands it with complementary context to efficiently pr…
cs.CV2024
ImagePiece: Content-aware Re-tokenization for Efficient Image Recognition
Seungdong Yoa, Seungjun Lee, Hyeseung Cho +2
Vision Transformers (ViTs) have achieved remarkable success in various computer vision tasks. However, ViTs have a huge computational cost due to their inherent reliance on multi-h…