search

RenYurui / Global-Flow-Local-Attention

The source code for paper "Deep Image Spatial Transformation for Person Image Generation"
https://renyurui.github.io/GFLA-web
Other
568 stars 84 forks source link

How is the key point detected #72

Closed code-conquer closed 3 years ago

code-conquer commented 3 years ago

Hello, how do you get information about human key points that is not mentioned in this paper or on this website?

RenYurui commented 3 years ago

Hi, I think there are some instructions about how to download the key points. For the pose image generation task, we use the key points provided by https://github.com/tengteng95/Pose-Transfer. (extracted by Openpose) For the animation task, the key points are extracted with the alphapose.