深度学习
海底小星星
这个作者很懒,什么都没留下…
展开
-
卷积神经网络——输入层、卷积层、激活函数、池化层、全连接层
卷积神经网络(CNN)由输入层、卷积层、激活函数、池化层、全连接层组成,即INPUT(输入层)-CONV(卷积层)-RELU(激活函数)-POOL(池化层)-FC(全连接层)卷积层用它来进行特征提取,如下: 输入图像是32*32*3,3是它的深度(即R、G、B),卷积层是一个5*5*3的filter(感受野),这里注意:感受野的深度必须和输入图像的深度相同。通过一个filter与输入图像的卷积可以...转载 2018-05-04 10:31:20 · 1527 阅读 · 0 评论 -
近期资料汇总
https://github.com/amusi/awesome-object-detection https://blog.csdn.net/xyy19920105/article/details/50817725 https://blog.csdn.net/qq_26898461/article/details/51262267 https://zhuanlan.zhihu.com/p/...原创 2019-03-10 22:39:14 · 145 阅读 · 0 评论 -
FPPI评价方式相关
https://blog.csdn.net/Bruce_0712/article/details/78462880 https://blog.csdn.net/u011534057/article/details/51734089原创 2019-03-19 18:38:37 · 498 阅读 · 0 评论 -
yolo相关资料
https://blog.csdn.net/guleileo/article/details/80581858 https://blog.csdn.net/m0_37192554/article/details/81092514 https://blog.csdn.net/zhangjunp3/article/details/80421814 https://blog.csdn.net/li...转载 2019-03-12 11:03:25 · 123 阅读 · 0 评论 -
目标检测相关论文
yolo_1:You Only Look Once:Unified, Real-Time Object Detection yolo9000:YOLO9000:Better, Faster, Stronger yolov3:YOLOv3: An Incremental Improvement RCNN:Rich feature hierarchies for accurate object det...原创 2019-03-16 17:05:31 · 193 阅读 · 0 评论 -
RCNN系列相关资料
https://www.cnblogs.com/dudumiaomiao/p/6560841.html https://blog.csdn.net/shenxiaolu1984/article/details/51152614 https://blog.csdn.net/shenxiaolu1984/article/details/51066975 https://blog.csdn.net...转载 2019-03-12 14:40:36 · 116 阅读 · 0 评论 -
Region Proposal by Guided Anchoring 阅读笔记
https://baijiahao.baidu.com/s?id=1626682069046284617&wfr=spider&for=pc https://blog.csdn.net/qq_37014750/article/details/86499652 https://blog.csdn.net/zbzcDZF/article/details/89480188原创 2019-07-01 20:00:10 · 315 阅读 · 0 评论 -
COCO数据集标注格式介绍
https://blog.csdn.net/scut_salmon/article/details/88252363 https://blog.csdn.net/yeyang911/article/details/78675942原创 2019-10-09 18:55:44 · 415 阅读 · 0 评论