PMYNet: A Light-weight Network Base on Image Channel Separation for Stereo Matching
Download as PDF
DOI: 10.23977/CNCI2020081
Author(s)
Yang Chen and Zongqing Lu
Corresponding Author
Yang Chen
ABSTRACT
With the continuous development of neural network research, the network has higher and higher requirements for memory and computational power, because the network has been designed to be more and more complex with tons of parameters. We compose a light-weight network with a pyramid module of YUV channel and attention block, which made a reasonable trade-off between the complexity of the network and the accuracy of the results.
KEYWORDS
Stereo matching; color; CNN; light-weight