Closed yukichou closed 1 year ago
The lib/data/dataset_action.py
has a coco2h36m
function that accepts a numpy ndarray of keypoints and does a rough conversion to the human3.6m format. I think I've seen other projects take a similar approach.
The
lib/data/dataset_action.py
has acoco2h36m
function that accepts a numpy ndarray of keypoints and does a rough conversion to the human3.6m format. I think I've seen other projects take a similar approach.
Thank you so much ! I will check this out as soon as possible.
@yukichou @Serdnad pose track and coco ,human3.6 ,AMASS ,there are all have different keypoints style, how to get unified keypoints ?
from the image above, it tend to be human3.6m keypoints format.
We convert the keypoints to H36M formats and the conversion code can be found in lib/data
.
I get the coco 17 key-points or any other key-point format of my own custom data, and I know I should convert the coco format to human3.6, but how? The definition between coco and human 3.6 is different especially for the body. Is there any way to convert the format between these datasets?