weixin_49636863的博客

原创 cpu ASRCF自用

https://blog.youkuaiyun.com/qq_45591903/article/details/108322699?spm=1001.2101.3001.6650.7&utm_medium=distribute.pc_relevant.none-task-blog-2%7Edefault%7EBlogCommendFromBaidu%7Edefault-7.no_search_link&depth_1-utm_source=distribute.pc_relevant.none-task

原创 RESNET50相关自用

https://blog.youkuaiyun.com/Enjolras_fuu/article/details/53710758?locationNum=5&fps=1https://www.cnblogs.com/yymn/articles/8666321.htmlhttps://www.jianshu.com/p/dd65cea075f6https://blog.youkuaiyun.com/SMF0504/article/details/90716694https://blog.youkuaiyun.com/qq_20

原创 分层特征

一https://blog.youkuaiyun.com/dcrmg/article/details/81255498下载keras下载权重每层特征提取二https://blog.youkuaiyun.com/Geek_of_优快云/article/details/84343971?utm_medium=distribute.pc_relevant.none-task-blog-2%7Edefault%7EBlogCommendFromMachineLearnPai2%7Edefault-1.control&a

原创 cf2

参考https://blog.youkuaiyun.com/crazyice521/article/details/65935753?utm_medium=distribute.pc_relevant.none-task-blog-2%7Edefault%7EBlogCommendFromBaidu%7Edefault-6.control&dist_request_id=&depth_1-utm_source=distribute.pc_relevant.none-task-blog-2%7Edefa

原创 DL CF

DLhttps://blog.youkuaiyun.com/qq_19329785/article/details/84504722https://blog.youkuaiyun.com/errors_in_life/article/details/65950699CF推导https://blog.youkuaiyun.com/sgfmby1994/article/details/68490903https://www.cnblogs.com/YiXiaoZhou/p/5925019.html数学https://blog.cs

原创 linux压缩解压

tar -cvf file2.tar checkpointscd 到checkpoints所在位置，将checkpoints文件夹压缩成file2.tarrar x baby.rar解压

原创 在自己的数据集上实现CycleGAN（pytorch linux）

代码https://github.com/junyanz/pytorch-CycleGAN-and-pix2pix参考文章https://blog.youkuaiyun.com/Gavinmiaoc/article/details/805855311.将数据裁成256*256。2.在datasets目录下新建文件夹label2image，在label2image下新建 trainA 和，trainB，testA，testB，A是你要训练的，B是你想生成的。(把图片放进去，不用成对)3.在你的pytorch下

原创 抠图 自用

https://blog.youkuaiyun.com/my_name_is_learn/article/details/103869155?utm_medium=distribute.pc_relevant.none-task-blog-BlogCommendFromMachineLearnPai2-3.control&dist_request_id=1328592.8641.16147369546321909&depth_1-utm_source=distribute.pc_relevant.non

原创 做图

错误No module named cityscapesscripts参考https://blog.youkuaiyun.com/m0_37833297/article/details/106873010https://blog.youkuaiyun.com/wang27623056/article/details/106631196/?utm_medium=distribute.pc_relevant.none-task-blog-baidujs-2%20https://www.cnblogs.com/leviatan/

原创 将自己的图片做成cityscape格式（自用）

1.labelme分割图片2.将json转label问题ModuleNotFoundError: No module named ‘labelme’解决pip install labelme==3.20.0问题module ‘labelme.utils’ has no attribute ‘draw_label’参考https://blog.youkuaiyun.com/qq_43571311/article/details/108490909解决因为labelme版本问题 少draw_labe

原创 远程服务器 Linux 用cityscape训练DeepLabv3模型（Pytorch版）并用图像测试

参考https://blog.youkuaiyun.com/qq_45389690/article/details/111591713?utm_medium=distribute.pc_relevant_download.none-task-blog-baidujs-2.nonecase&depth_1-utm_source=distribute.pc_relevant_download.none-task-blog-baidujs-2.nonecasehttps://blog.youkuaiyun.com/weixi

原创 自制语义分割数据集 出现的错误问题

参考https://blog.youkuaiyun.com/wjinjie/article/details/106735141出现两个问题1.cd进入json时报错将这句代码手打进去 就好了 labelme_json_to_dataset 0002.json2.生成的文件中没有info.yaml参考https://blog.youkuaiyun.com/winter616/article/details/104426111得知是labelme版本的问题 新版本中代码没有生成info.yaml文件解决办法 在lab

原创 pytorch gpu改cpu

https://www.cnblogs.com/gaona666/p/12357282.htmlhttps://ptorch.com/news/71.htmlhttps://blog.youkuaiyun.com/weixin_42274540/article/details/80653450https://blog.youkuaiyun.com/a8039974/article/details/109622037https://blog.youkuaiyun.com/weixin_36670529/article/details/1

原创 win10-pytorch-cpu

因为显卡不可，所以只能cpu1.首先 在anaconda prompt(anaconda3)创建一个环境(base) C:\Users\YEYE>conda create -n pytorch2.打开环境conda activate pytorch3.在pytorch官网查看合适版本https://pytorch.org/get-started/locally/conda install pytorch torchvision torchaudio cpuonly -c pytorc

原创 UAV123测试自己的跟踪算法

插入链接与图片链接:https://blog.youkuaiyun.com/qq_17783559/article/details/106983082https://blog.youkuaiyun.com/qq_17783559/article/details/82081604https://blog.youkuaiyun.com/yinxian9019/article/details/107055601/?utm_medium=distribute.pc_relevant.none-task-blog-title-7&spm=

原创 matlab2017a和vs2015

原创 ‘Tensor‘ object has no attribute ‘numpy‘问题解决

问题：tensorflow的版本问题，没开启eager_execution解决：import tensorflow as tf后加tf.enable_eager_execution()！！！！一定要 new console后才好使！！！！！参考添加链接描述

原创 No module named ‘imgaug‘问题解决

No module named 'imgaug’问题解决·刚开始 pip install imgaug安装成功了，但跑代码时依然报错，仍旧没有这个模块·后来 conda install -c conda-forge imgaug成功！！！·查了一下原因，conda会检查当前环境下所有包之间的关系，pip不管，导致用pip安装时，即使安装成功，也无法正常使用。...