MartinXM / GAP

official implementation for Language Supervised Training for Skeleton-based Action Recognition
Apache License 2.0
99 stars 11 forks source link

2D data set use case possibility? #1

Closed mukeshnarendran7 closed 1 year ago

mukeshnarendran7 commented 1 year ago

Hi, thanks for sharing your work and I am interested in trying to another application. I am aware that the dataset is based on 3D co-ordinates but I have a 2D pose dataset. Will it be possible to incorporate that data to train this model using transfer learning? If yes, how can i do so?Thanks

MartinXM commented 1 year ago

2D pose to 3D joints transfer learning is an interesting topic. However, the 2D and 3D datasets have different input dimensions, maybe you need to tackle that first.