Badour AlBahar

I am also using the pretrained model and I get "okay" results. but the ssim is 0.6 and not 0.7 as reported not sure why. Here are some of the...

Do you represent the coordinates as 18 channels where the (x, y) joint location is set to 1 and everything else is set to 0?

I have the same issue. I, however, am not changing the model. I am using pytorch 1.2.0.

How do we generate the segmentation files? Which human parser do you use?

Thank you very much for the resources. In [here](https://github.com/foamliu/Look-Into-Person), there are 20 classes for the human parts, how did you get the 8 segmented parts? Edit: you mention in the...

Hi Erika, The “weird” part for me is the pink color. I thought it should look blue/green. I was also wondering how you trained this model. What was the loss...

Thank you for the information Erika. Can you share the code for training the kp2dp model?

I have the same issue too. ``` PyTorch version: 1.6.0 CUDA used to build PyTorch: 10.2 OS: CentOS Linux release 7.9.2009 (Core) (x86_64) GCC version: (GCC) 5.5.0 Libc version: glibc-2.17...

I am basically following the steps of the [rendering pipeline](https://github.com/lingjie0206/Neural_Actor_Main_Code/blob/master/docs/rendering_pipeline.md). This is the output of "STEP 3: generate video given camera poses".

``` >>> import torch; print(torch.__version__, torch.version.cuda) 1.6.0 10.2 >>> import torchvision; print(torchvision.__version__) 0.7.0 ```