Viet-Anh on Software Logo

What is: ShuffleNet v2?

SourceShuffleNet V2: Practical Guidelines for Efficient CNN Architecture Design
Year2000
Data SourceCC BY-SA - https://paperswithcode.com

ShuffleNet v2 is a convolutional neural network optimized for a direct metric (speed) rather than indirect metrics like FLOPs. It builds upon ShuffleNet v1, which utilised pointwise group convolutions, bottleneck-like structures, and a channel shuffle operation. Differences are shown in the Figure to the right, including a new channel split operation and moving the channel shuffle operation further down the block.