Dynamic Convolution: Attention over Convolution Kernels

these kernels are aggregated in a non-linear way via atten-tion. By simply using dynamic convolution for the state-of-the-art architecture MobileNetV3-Small, the top-1 accuracy of ImageNet classiﬁcation is boosted by 2.9% with only 4% additional FLOPs and 2.9 AP gain is achieved on COCO keypoint detection. 1. Introduction

Tags:

Information