arrow
返回

Image content-dependent steerable kernels

delete2021-04-26
delete1
PRE
AI
X
Xiang Ye
H
Heng Wang
Y
Yong Li *
DOI:10.1007/s00371-021-02128-zdelete
delete原文链接
delete原文求助
delete分享
delete收藏
摘要

摘要

En 中文
Attention mechanism plays an essential role in many tasks such as image classification, object detection, and instance segmentation. However, existing methods typically assigned attention weights to feature maps of the previous layer. The kernels in current layer remained static during the inference stage. To explicitly model the dependency of individual kernel weights on image content at the inference stage, this work proposed attention weight block (AWB) that converts kernels to be steerable to the content in a test image. Specifically, AWB computes a set of on-the-fly coefficients according to the feature maps of the previous layer and applies the coefficients to the kernels in current layers, which makes them steerable. AWB kernels emphasize or suppress the weights of certain kernels depending on the content of input samples and hence significantly improve the feature representation ability of deep neural networks. The proposed AWB is evaluated on various datasets, and experimental results show that steerable kernels in AWB outperformed the state-of-the-art attention approaches when embedded in the architecture for classification, object detection, and semantic segmentation tasks. It outperforms ECA by 1.1% and 1.0% on CIFAR-100 and Tiny ImageNet datasets, respectively, for image classification task; outperforms CornerNet-Lite by 1.5% on COCO2017 dataset for object detection task; and outperforms FCN8s by 1.2% on SBUshadow dataset for semantic segmentation task.
AI总结

AI总结

对已上传原文的论文进行重点信息的提取,主要内容包括:简要概述、研究摘要、背景介绍、关键亮点、图文解析、展望与总结。

期刊

Visual Computer 封面图
Visual Computer
IF:
2.9
论文数:
4.6K
被引数:
6.5K

机构

B
beijing university of posts & telecommunications
学者数:
1.4W
论文数: 1.2W
被引数: 9
引用论文

引用论文

Effects of eccentricity defect on the nonlinear dynamic behavior of the mechanism clutch-helical two stage gear
err2011-07-01
err0
PREAI
errLassâad Walha; Yassine Driss; Mohamed Taoufik Khabou; Tahar Fakhfakh; Mohamed Haddar
err分享
err收藏
Multi-level progressive parallel attention guided salient object detection for RGB-D images
err2020-02-18
err15
PREAI
errLiu, Zhengyi; Duan, Quntao; Shi, Song; Zhao, Peng
err分享
err收藏
Synthesis of n-type semiconducting diamond film using diphosphorus pentaoxide as the doping source以五氧化二磷为掺杂源合成n型半导体金刚石膜
err1990-10-01
err0
PREAI
errKen Okano; Hideo Kiyota; Tatsuya Iwasaki; Yoshitaka Nakamura; Yukio Akiba; Tateki Kurosu; Masamori Iida; Terutaro Nakamura
err分享
err收藏
err分享
err收藏
学者 查看更多内容