New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Validate the model #83
Comments
|
Suppose I have a model file of an aircraft to use
ocnn_M40_5.caffemodel to test his accuracy, how can I use the command
line, thank you
抱歉打扰您,由于我是初学者,想问您如何加载自己的.points档去让已训练好的模型分辨,command
line如何下,因为在网上查询的多是图片辨识,都要有个均值档跟deploy.prototxt才能进行分辨,谢谢您我可能说的不够明确,我举个例子,假设我有一个飞机的模型档要用ocnn_M40_5.caffemodel,来测试他的准确性,那我要如何下command
line,谢谢
Peng-Shuai Wang <notifications@github.com>於 2020年3月25日 週三,下午7:48寫道:
…
1.
Run the following command to test:
python run_cls.py configs/cls_octree_test.yaml
2.
For the problem Cannot copy param 0 weights from layer 'ip2'; shape
mismatch. Source param shape is 40 128 ... , the reason is that when
training on ModelNet40, the dimension of the last FC is 40x128 for the
40-category classification, but the FC of your network is 5x128 for the
4-category classification. So you have to rename the layer's parameters to
train it from scratch.
3.
测试的时候数据处理跟训练一样,需要将.off转换为.points或者.octree
—
You are receiving this because you authored the thread.
Reply to this email directly, view it on GitHub
<#83 (comment)>, or
unsubscribe
<https://github.com/notifications/unsubscribe-auth/AE4TAUCME6VA2HXJN35JJ4DRJHVRFANCNFSM4LTH76JA>
.
|
您好
不好意思又打扰您
在阅读您的文献时有些问题,不知道您方不方便为我解惑,在o-cnn中,不同的模型以八叉树做切割,那假设今天有两个同样是飞机模型,有可能切
下去的相同位置在这两个模型上是不同位置,想请问这样要如何辨识他是飞机模型,假设一个飞机头的座标与另一个正好相反之类的情况。
谢谢您
Peng-Shuai Wang <notifications@github.com>於 2020年3月26日 週四,下午9:33寫道:
… Closed #83 <#83>.
—
You are receiving this because you authored the thread.
Reply to this email directly, view it on GitHub
<#83 (comment)>, or
unsubscribe
<https://github.com/notifications/unsubscribe-auth/AE4TAUBBB4QAENKE7VTPON3RJNKRXANCNFSM4LTH76JA>
.
|
您好 |
@chenteng1998 你运行的命令缺少关键字
|
谢谢!!我的问题已经解决 |
Hello, would you like to ask how do you perform a verification test on the trained model and what kind of file is the test, is it .off or .points, or an octree file?
I want to use your model for pre-train, but I have the following problem:
Cannot copy param 0 weights from layer 'ip2'; shape mismatch. Source param shape is 40 128 (5120); target param shape is 5 128 (640). To learn this layer's parameters from scratch rather than copying from a saved net, rename the layer.
你好,想请问您如何对训练好的模型进行验证测试,测试的档案类别为何,是.off还是.points,或是octree档
The text was updated successfully, but these errors were encountered: