Different input output size

garyzhao / SemGCN

The Pytorch implementation for "Semantic Graph Convolutional Networks for 3D Human Pose Regression" (CVPR 2019).

https://arxiv.org/abs/1904.03345

Apache License 2.0

467 stars 78 forks source link

Different input output size #53

Open funmonty opened 2 years ago

funmonty commented 2 years ago

Hi,

I can see that the repo accepts adjacency matrix shape and input-output shape to be the same NxN. N keypoints in input and N in the output. Is it possible to have it as MxN? Meaning, I want to derive N 3d keypoints from M 2d keypoints in my usecase.