SUPER-MARIO
SUPER-MARIO
@xbpeng Thanks for your patience. However, for me, it seems the velocities are too small. And the actions and rewards are easily to get nan value. I think the initial...
@Zju-George I've converted the result motion already so that I can use it to train policy, I have obtained the root translation myself, but I think it's not good enough,...
Yes, I've implemented ASI, it converges much faster than RSI, it obtains good results on some easy clips(those shows in the paper), but it still fails if the motion result...
@chingswy Thanks, and how to match data? I see you load a pre-computed matrix, but how can I get this?
And how about the code to compute affinity matrix and the code of the optimization process with low rank constraint
@DongJT1996 Looking forward.
Hi, thanks for your comment, seems you're right.
@ZoroDerVonCodier @manuelschmidt Thanks for both of you. I'd be appreciated if you send a pull request to fix this issue.
@ZoroDerVonCodier, Thanks again for your attention. Do you have any idea to solve this?
@makroid The unpooling layer is to reproduce the paper **Visualizing and Understanding Convolutional Networks**, it can only be used for visualization, so it cannot be used to train the network....