knight

Consult the problems about the function ChannelizeImage() and preconditioner.F_fft * model_precond.F_fft

1

Hi, @jonbarron , thanks for sharing the ffcc project. I am researching the ffcc project and have met two problems as follows. **First, I can not understand why use the...

How to understand the meaning of the conv operations in the function "RenderHistorgramGaussian"?

2

hi， @jonbarron, thanks for sharing the codes of the ffcc. I can't understand the meaning of the conv operations in the function "RenderHistorgramGaussian". The code is as follow, ``` %...

When params.DEEP.ENABLED is true, there is problem about the param.DEEP.FEATURE_FILENAME_TAGS.

1

hi, When i set the params.DEEP.ENABLED true, it apperas the bug about the `length(params.DEEP.FEATURE_FILENAME_TAGS)`. The define of the params.DEEP.FEATURE_FILENAME_TAGS is 'feature'.

非常感谢你提供的训练联合贝叶斯和测试LFW的代码，困惑了好久的一个问题，想请教您一下： 我采用的Webface的features训练联合贝叶斯时得到联合贝叶斯的四个模型参数，接下来我用LFW的九分数据（5400对）按照您设置阈值的方式，计算出阈值，然后用另外一份（600对）数据进行测试，交叉验证的方法得到一个平均精度,不知道我这样的思路对不对？ 在设置阈值时采用 Webface 数据来设置阈值还是根据LFW来设置阈值？ 我看有些开源的代码直接设置的一个常数，但是不知道常数是怎么设置的？

``` z = z + convn(net.layers{l + 1}.d{j}, rot180(net.layers{l + 1}.k{i}{j}), 'full'); net.layers{l}.dk{i}{j} = convn(flipall(net.layers{l - 1}.a{i}), net.layers{l}.d{j}, 'valid') / size(net.layers{l}.d{j}, 3); ``` how to understand rot180 and 'full'? how...

[BUG] ModuleNotFoundError: No module named 'onnxsim.onnxsim_cpp2py_export'

2

**Describe the bug** 对 onnx 模型进行简化时报 ModuleNotFoundError: No module named 'onnxsim.onnxsim_cpp2py_export'： 在 Conda 虚拟环境下已安装 onnxsim 版本号是 0.4.36， 而且 import onnxsim 下是能够找到 onnxsim_cpp2py_export， 但是不知道是什么原因一直报上述错误，求帮助。 **Model** To reproduce the problem, please post...

How to solve the problem of model convergence? The model is trained on the training data 16K(default parmameter) for 200 epochs, and then tested on the same data, but the...