Efficient Vision Transformers via Token Merging with Head-wise Attention Correction

Publication Information

Title

Japanese:	Efficient Vision Transformers via Token Merging with Head-wise Attention Correction
English:	Efficient Vision Transformers via Token Merging with Head-wise Attention Correction

Author

Japanese:	Ichikawa, Yuki, Motomura, Masato, Van Chu, Thiem, 藤木大地.
English:	Ichikawa, Yuki, Motomura, Masato, Van Chu, Thiem, Daichi Fujiki.

Language

English

Journal/Book name

Japanese:	Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision
English:	Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision

Volume, Number, Page

pp. 3908-3917

Published date

Mar. 6, 2026

Publisher

Japanese:
English:

Conference name

Japanese:
English:

Conference site

Japanese:
English: