DynamicViT: Efficient Vision Transformers with Dynamic Token Sparsification.

Yongming Rao Wenliang Zhao Benlin Liu Jiwen Lu Jie Zhou Cho-Jui Hsieh

Published in: CoRR (2021)

Keyphrases

real time
computer vision
database
image processing
vision system
real world
preprocessing
visual perception
high efficiency
computationally efficient
dynamic environments
relational databases
lower bound
object recognition
search algorithm
data structure
bayesian networks
three dimensional
information retrieval