深度学习论文阅读
yaoling-xumi13
这个作者很懒,什么都没留下…
展开
-
Stacked Hourglass Networks for Human Pose Estimation
一.概述1.paper: Stacked Hourglass Networks for Human Pose Estimation2.code: http://www-personal.umich.edu/~alnewell/pose3.Time: 2016034.task: human pose estimation5.eval_dataset: FLIC and MPII bench...原创 2020-03-29 20:20:31 · 450 阅读 · 0 评论 -
Deeplab v3+论文笔记
Deeplab v3+Deeplab v3+的结构1) Spatial pyramid poolingencode multi-scale contextual information by probing the incoming features.2) encode-decoder structurecapture shaper object boundaries by gra...原创 2020-03-01 22:40:27 · 1155 阅读 · 1 评论 -
空洞卷积Atrous convolution和ASPP
1.Atrous Convolutionfrom: https://towardsdatascience.com/review-deeplabv1-deeplabv2-atrous-convolution-semantic-segmentation-b51c5fbde92dThe term “Atrous” indeed comes from French “à trous” meanin...原创 2020-02-29 22:44:00 · 1990 阅读 · 0 评论 -
论文阅读:Accurate, Large Minibatch SGD: Training ImageNet in 1 Hour
论文阅读:Accurate, Large Minibatch SGD: Training ImageNet in 1 Hour论文地址: https://arxiv.org/pdf/1706.02677背景1)larger networks and larger datasets need longer time for training; 解决方法:分布式同步SGD,将SGD minib...原创 2019-08-18 23:15:00 · 1355 阅读 · 0 评论 -
车道线-论文阅读: Learning Lightweight Lane Detection CNNs by Self Attention Distillation
ICCV2019code: https://github.com/cardwing/Codes-for-Lane-Detectionpaper: https://arxiv.org/abs/1908.00821Abstractpresent a novel knowledge distillation approach, i.e., Self Attention Distillatio...原创 2019-09-01 17:54:07 · 1830 阅读 · 0 评论 -
论文阅读- Paying More Attention to Attention: Improving the Performance of Convolutional Neural Networks
Paying More Attention to Attention: Improving the Performance of Convolutional Neural Networks via Attention Transfer论文地址:https://arxiv.org/abs/1612.03928github地址:https://github.com/szagoruyko/atten...原创 2019-09-01 23:34:15 · 5624 阅读 · 0 评论