2 papers
cs.CV2022
Self-Knowledge Distillation via Dropout
Hyoje Lee, Yeachan Park, Hyun Seo +1
To boost the performance, deep neural networks require deeper or wider network structures that involve massive computational and memory costs. To alleviate this issue, the self-kno…
cs.CV2019
Attention routing between capsules
Jaewoong Choi, Hyun Seo, Suii Im +1
In this paper, we propose a new capsule network architecture called Attention Routing CapsuleNet (AR CapsNet). We replace the dynamic routing and squash activation function of the…