How the model will work if I want the output with the same pose as the reference image?

[Mushahid2521]

I was wondering how I can modify the existing model to achieve the output with the same pose as reference image.

1. Will it reduce the inference time?
2. Will it reduce the model size?

[MotiBaadror]

Here are steps that help you do this.

1. Use open pose to estimate pose for the reference image and put that json file in the ~datasets/pose_coco folder
2. Go to demo/demo.txt file in that modify the name of keypoint.json to your json
   Now running demo.sh will give the output with reference pose into demo/forward folder

[Mushahid2521]

Thanks a lot for your reply.
So what I want is that if I don't output with a new pose I don't require the face model. So the face refinement step can be ignored I think.
So how to achieve this?

If I use same pose as reference then can I expect to get the exact face as reference if I follow your procedure?

[MotiBaadror]

Unfortunately that is not happening. I tried it.

…

On Tue, Sep 1, 2020, 12:34 PM Mushahidul Islam Shamim < ***@***.***> wrote: Thanks a lot for your reply. So what I want is that if I don't output with a new pose I don't require the face model. So the face refinement step can be ignored I think. So how to achieve this? If I use same pose as reference then can I expect to get the exact face as reference if I follow your procedure? — You are receiving this because you commented. Reply to this email directly, view it on GitHub <#29 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AG46Y6IMLQTUKRT6SPVPT4TSDSMG5ANCNFSM4QPQGSXA> .