Kaini J

Regarding the issue of pre-trained weights.

1

Hello, I am amazed by your outstanding work. However, I have a small question that puzzles me. Based on your code's parameter settings, the accuracy only reaches the second-best accuracy...

你好，请问Export coco labels data set v1 这一步，是默认在cpu上执行吗，一共会有多少个npz文件导出呀，我已经跑了一天多，不太确定还需多长时间

2

Export COCO labels data set

I have modified the backbone of the MagicPoint network and am unsure about the accuracy metrics after step 1) Training MagicPoint on Synthetic Shapes in the original network. Could you...

Hello, I am not sure if there is an issue with the way I generated the COCO training dataset. Previously, some comments mentioned that it only takes about 20 minutes...

MillionAID数据集中的真值场景分类编号

1

最近想把场景分类作为一个语义信息提取的前置处理过程，目的是想实现对单张任意影像的类别预测。但是遇到的问题还是millionAID的真值标签信息缺失。 目前可以通过网络的前向传播输出一张影像的类别号，但是不知道对应的类别具体是哪个，查看了之前的issue发现训练数据采用的是随机抽取的方式，这种情况下不清楚预训练网络的输出针对我们自己定义的label是否仍然适用？ 欢迎各位大佬指导我这个小菜鸡@_@

Runtime issues related to GPU.

1

Has anyone encountered this issue? An error occurs when running the following command “python examples/demo_how.py eval examples/params/eccv20/eval_how_r50-_2000.yml -e official_how_r50-_2000” ，because the hardware does not support CUDA 10.1, making it impossible...