ShiftAddViT: Mixture of Multiplication Primitives Towards Efficient Vision Transformer.

Haoran You Huihong Shi Yipin Guo Yingyan Lin

Published in: CoRR (2023)

Keyphrases

image processing
knowledge base
low level
high level
vision system
information retrieval
image segmentation
multiscale
fault diagnosis
machine learning
database systems
data model
mobile robot
lightweight
building blocks