PolyNet: A Pursuit of Structural Diversity in Very Deep Networks

Xingcheng ZhangZhizhong LiChen Change LoyDahua Lin

   Papers with code   Abstract  PDF

A number of studies have shown that increasing the depth or width of convolutional networks is a rewarding approach to improve the performance of image recognition. In our study, however, we observed difficulties along both directions... (read more)

Benchmarked Models

RANK
MODEL
REPO
CODE RESULT
PAPER RESULT
ε-REPRODUCED
BUILD
1
PolyNet
80.9%
81.3%