sankexin
sankexin
You only need to install detectron2. We recommend the Pre-Built Detectron2 (Linux only) version with pytorch 1.5. I use the Pre-Built Detectron2 with CUDA 10.1 and pytorch 1.5 and you...
well, I'll try it myself.I found that some target detection is not good, mainly because of the low similarity between these targets and the targets in the pre training, the...
把权重文件放到ncnn目录（根目录下）
报这个错： param is too old, please regenerate! layer load_param 258 ReduceMean_539 failed param is too old, please regenerate! layer load_param 260 ReduceMean_541 failed Segmentation fault (core dumped) 有如何转换的代码吗？
> param 在头部写进去报错，param里加哪里？怎么加的示例有吗？
> > > param > > > 在头部写进去报错，param里加哪里？怎么加的示例有吗？ > > Reduction ReduceMean_611 1 1 src2_splitncnn_2 1088 0=3 1=0 -23303=1,1 4=1 5=1 类似这种 在后边加个5=1 还是报错： Segmentation fault (core dumped) 模型可以加载了，但获取fgr和pha时就报错这个错了，看起来像是获取到的bin里的权重是错的，你有rvm mobilenetv3转换成ncnn的代码可以我自己转换试试效果吗？
这是来自QQ邮箱的假期自动回复邮件。 您好，我最近正在休假中，无法亲自回复您的邮件。我将在假期结束后，尽快给您回复。
这是来自QQ邮箱的假期自动回复邮件。 您好，我最近正在休假中，无法亲自回复您的邮件。我将在假期结束后，尽快给您回复。
怎么让网络获取到正确的权重呢@nihui
> how to contact the author for the dataset? I cannot find an email address in the repo. Also the paper has not been released... @sanshibayuan Have you received the...