PokerNet:Expanding Features Cheaply via Depthwise Convolutions

来源 :International Journal of Automation and Computing | 被引量 : 1次 | 上传用户:qwerasd1234
下载到本地 , 更方便阅读
声明 : 本文档内容版权归属内容提供方 , 如果您对本文有版权争议 , 可与客服联系进行内容授权或下架
论文部分内容阅读
Pointwise convolution is usually utilized to expand or squeeze features in modern lightweight deep models. However, it takes up most of the overall computational cost(usually more than 90%). This paper proposes a novel Poker module to expand features b
其他文献
以鼓楼大街车站基坑工程施工为背景,主要介绍了地铁车站周围环境与车站基坑围护结构,采用地下连续墙围护体系作业工艺进行施工,提出以成槽加固与坑内土体加固、导墙施工、泥
针对柳城特长隧道穿越高地温地区施工存在温度高、湿度大、高温段落长的特点,系统分析了高地温隧道衬砌的受力特征,提出了高地温隧道隔热和降温综合控制措施,解决了隧道内施
Audio-visual learning, aimed at exploiting the relationship between audio and visual modalities, has drawn considerable attention since deep learning started to be used successfully. Researchers tend