SUPER-MARIO

Model blending using only 1_F is better than using all three networks in level1.

1

When I test level1, I found that using 1_F only is better than using 1_F + 1_EN + 1_NM. By the way, the loss of 1_EN and 1_NM is also...

About the modified caffe

1

You said that you have modified caffe source code so it would log every test iter, would you mind sharing your code? Thanks.

请问代码里的initial_guess和getsimilarTransform是什么意思？谢谢

6

如题，谢谢，没有太懂。

The training process is not the same as the paper.

1

The paper use as training method called "Alternating Training", but the code just train the model end to end.

It's not easy to use openpose to generate the input data that can be used for my own video.

What does the mask label look like?

2

From the following code, we can compute the mask loss, but I cannot understand what does the mask label look like, I've read through the code for loading data, I...

Since OneDrive is blocked....

How you preprocess the data?

1

Hi, I want to know how you preprocess the data into rec file, what data augmentation has been used?

About GetMeanShape

11

In the function GetMeanShape, `halfsize = int((ymean - ymin) * 2.0)`, I think it's not proper to `*2`.

How about the speed?

1

Hi, have you tested the speed of the model?